Revert "[BugFix] Fix 'GPUARWorker' attribute error (#5478)"

This reverts commit f4640d9a2e.

.buildkite/amd/scripts/bootstrap-amd-omni.sh

@@ -6,6 +6,8 @@
 
 set -euo pipefail
 
+source .buildkite/common/scripts/resolve_skip_ci.sh
+
 if [[ -z "${RUN_ALL:-}" ]]; then
     RUN_ALL=0
 fi
@@ -22,10 +24,6 @@ if [[ -z "${AMD_MIRROR_HW:-}" ]]; then
     AMD_MIRROR_HW="amdproduction"
 fi
 
-if [[ -z "${DOCS_ONLY_DISABLE:-}" ]]; then
-    DOCS_ONLY_DISABLE=0
-fi
-
 fail_fast() {
     DISABLE_LABEL="ci-no-fail-fast"
     # If BUILDKITE_PULL_REQUEST != "false", then we check the PR labels using curl and jq
@@ -56,88 +54,6 @@ check_run_all_label() {
     fi
 }
 
-is_docs_only_change() {
-    local file_path
-    local has_any=0
-
-    while IFS= read -r file_path; do
-        [[ -z "${file_path}" ]] && continue
-        has_any=1
-
-        if [[ "${file_path}" == docs/* ]]; then
-            continue
-        fi
-        if [[ "${file_path}" == *.md ]]; then
-            continue
-        fi
-        if [[ "${file_path}" == "mkdocs.yml" ]]; then
-            continue
-        fi
-        return 1
-    done
-
-    [[ "${has_any}" -eq 1 ]]
-}
-
-resolve_skip_ci() {
-    local is_pr_build=0
-    local files
-    local base_branch base_ref
-
-    if [[ "${BUILDKITE_PULL_REQUEST:-false}" != "false" && -n "${BUILDKITE_PULL_REQUEST:-}" ]]; then
-        is_pr_build=1
-    fi
-
-    if [[ "${is_pr_build}" -eq 1 ]]; then
-        base_branch="${BUILDKITE_PULL_REQUEST_BASE_BRANCH:-main}"
-        if ! git rev-parse --verify "origin/${base_branch}" >/dev/null 2>&1; then
-            echo "resolve_skip_ci: origin/${base_branch} not found locally; trying fetch" >&2
-            git fetch --depth=200 origin "${base_branch}" >/dev/null 2>&1 || true
-        fi
-
-        base_ref=""
-        if git rev-parse --verify "origin/${base_branch}" >/dev/null 2>&1; then
-            base_ref="origin/${base_branch}"
-        elif git rev-parse --verify "${base_branch}" >/dev/null 2>&1; then
-            base_ref="${base_branch}"
-        else
-            echo "resolve_skip_ci: cannot resolve PR base ${base_branch}; skip-ci=0" >&2
-            echo -n 0
-            return 0
-        fi
-
-        if ! files="$(git diff --name-only "${base_ref}...${BUILDKITE_COMMIT}" 2>/dev/null)"; then
-            echo "resolve_skip_ci: failed to compute PR changed files; skip-ci=0" >&2
-            echo -n 0
-            return 0
-        fi
-    elif [[ "${BUILDKITE_BRANCH:-}" == "main" ]]; then
-        if ! git rev-parse --verify "${BUILDKITE_COMMIT}^" >/dev/null 2>&1; then
-            echo "resolve_skip_ci: commit has no parent on main; skip-ci=0" >&2
-            echo -n 0
-            return 0
-        fi
-        if ! files="$(git diff --name-only "${BUILDKITE_COMMIT}^..${BUILDKITE_COMMIT}" 2>/dev/null)"; then
-            echo "resolve_skip_ci: failed to compute main changed files; skip-ci=0" >&2
-            echo -n 0
-            return 0
-        fi
-    else
-        echo "resolve_skip_ci: not PR/main build; skip-ci=0" >&2
-        echo -n 0
-        return 0
-    fi
-
-    if is_docs_only_change <<< "${files}"; then
-        echo "resolve_skip_ci: docs-only change detected; skip-ci=1" >&2
-        echo -n 1
-        return 0
-    fi
-
-    echo "resolve_skip_ci: non-doc changes detected; skip-ci=0" >&2
-    echo -n 0
-}
-
 if [[ -z "${COV_ENABLED:-}" ]]; then
     COV_ENABLED=0
 fi
@@ -154,14 +70,14 @@ upload_pipeline() {
     fi
 
     # Use local template file for vllm-omni
-    cp .buildkite/test-template-amd-omni.j2 .buildkite/test-template.j2
+    cp .buildkite/amd/test-template-amd-omni.j2 .buildkite/amd/test-template.j2
 
 
     # (WIP) Use pipeline generator instead of jinja template
-    if [ -e ".buildkite/pipeline_generator/pipeline_generator.py" ]; then
+    if [ -e ".buildkite/amd/pipeline_generator/pipeline_generator.py" ]; then
         python -m pip install click pydantic
-        python .buildkite/pipeline_generator/pipeline_generator.py --run_all=$RUN_ALL --list_file_diff="$LIST_FILE_DIFF" --nightly="$NIGHTLY" --mirror_hw="$AMD_MIRROR_HW"
-        buildkite-agent pipeline upload .buildkite/pipeline.yaml
+        python .buildkite/amd/pipeline_generator/pipeline_generator.py --run_all=$RUN_ALL --list_file_diff="$LIST_FILE_DIFF" --nightly="$NIGHTLY" --mirror_hw="$AMD_MIRROR_HW"
+        buildkite-agent pipeline upload .buildkite/amd/pipeline.yaml
         exit 0
     fi
     echo "List file diff: $LIST_FILE_DIFF"
@@ -171,7 +87,7 @@ upload_pipeline() {
 
     FAIL_FAST=$(fail_fast)
 
-    cd .buildkite
+    cd .buildkite/amd
 
     # Select test definition file: merge suite for main, ready suite for PRs.
     # For debugging, DEBUG_TEST_YAML can override the selection — accepts
@@ -182,7 +98,7 @@ upload_pipeline() {
                 TEST_YAML="test-amd-merge.yml"
                 ;;
             ready)
-                TEST_YAML="test-amd-ready.yaml"
+                TEST_YAML="test-amd-ready.yml"
                 ;;
             *)
                 echo "ERROR: DEBUG_TEST_YAML must be 'merge' or 'ready', got '$DEBUG_TEST_YAML'" >&2
@@ -193,7 +109,7 @@ upload_pipeline() {
     elif [[ $BUILDKITE_BRANCH == "main" ]]; then
         TEST_YAML="test-amd-merge.yml"
     else
-        TEST_YAML="test-amd-ready.yaml"
+        TEST_YAML="test-amd-ready.yml"
     fi
 
     (
@@ -234,23 +150,13 @@ if [[ $BUILDKITE_BRANCH == "main" ]]; then
     file_diff=$(get_diff_main)
 fi
 
-# ----------------------------------------------------------------------
-# Early exit start: skip pipeline if conditions are met
-# ----------------------------------------------------------------------
-
-# Match CUDA Buildkite skip-ci behavior for docs/markdown-only changes.
-if [[ "${DOCS_ONLY_DISABLE}" != "1" ]]; then
-    SKIP_CI="$(resolve_skip_ci)"
-    if [[ "${SKIP_CI}" == "1" ]]; then
-        echo "[docs-only] Docs/markdown-only changes detected. Exiting before pipeline upload."
-        exit 0
-    fi
+# Early exit: unified skip-ci (docs / skip marks) and CI-yaml-only level targeting.
+if [[ $BUILDKITE_BRANCH == "main" ]]; then
+    gate_bootstrap_ci amd l3
+else
+    gate_bootstrap_ci amd l2
 fi
 
-# ----------------------------------------------------------------------
-# Early exit end
-# ----------------------------------------------------------------------
-
 patterns=(
     "docker/Dockerfile"
     "CMakeLists.txt"

.buildkite/amd/scripts/run-amd-test.sh

@@ -56,7 +56,7 @@ container_name="rocm_${BUILDKITE_COMMIT}_$(tr -dc A-Za-z0-9 < /dev/urandom | hea
 # # Install AWS CLI to authenticate to ECR Public Gallery to get higher rate limit for pulling images
 # sudo apt-get update && sudo apt-get install -y awscli
 ##  Use safe docker login helper to prevent race conditions
-# source "$(dirname "${BASH_SOURCE[0]}")/../docker_login_ecr_public.sh"
+# source "$(dirname "${BASH_SOURCE[0]}")/../../common/scripts/docker_login_ecr_public.sh"
 # safe_docker_login_ecr_public
 
 ## Pull the container from AMD Docker Hub

.buildkite/amd/test-template-amd-omni.j2

@@ -57,7 +57,7 @@ steps:
           {% endif %}
 {% set cmd_body = (step.command or (step.commands | join("\n"))) | trim %}
 {% set indented_cmd = cmd_body | replace("\n", "\n            ") %}
-        command: bash .buildkite/scripts/hardware_ci/run-amd-test.sh
+        command: bash .buildkite/amd/scripts/run-amd-test.sh
         env:
           DOCKER_BUILDKIT: "1"
 {% if step.env %}
@@ -104,7 +104,7 @@ steps:
       {% endif %}
 {% set cmd_body = (step.command or (step.commands | join("\n"))) | trim %}
 {% set indented_cmd = cmd_body | replace("\n", "\n        ") %}
-    command: bash .buildkite/scripts/hardware_ci/run-amd-test.sh
+    command: bash .buildkite/amd/scripts/run-amd-test.sh
     env:
       DOCKER_BUILDKIT: "1"
 {% if step.env %}

.buildkite/bootstrap-intel-omni.sh

@@ -1,51 +0,0 @@
-#!/bin/bash
-# vllm-omni Intel bootstrap
-# Uses static pipeline-intel.yaml for Intel XPU tests
-# Based on: bootstrap-amd-omni.sh
-
-set -euo pipefail
-
-if [[ -z "${DOCS_ONLY_DISABLE:-}" ]]; then
-    DOCS_ONLY_DISABLE=0
-fi
-
-upload_pipeline() {
-    echo "--- 🛠 Preparing Intel pipeline"
-    buildkite-agent pipeline upload .buildkite/pipeline-intel.yaml
-}
-
-get_diff() {
-    $(git add .)
-    echo $(git diff --name-only --diff-filter=ACMDR $(git merge-base origin/main HEAD))
-}
-
-get_diff_main() {
-    $(git add .)
-    echo $(git diff --name-only --diff-filter=ACMDR HEAD~1)
-}
-
-file_diff=$(get_diff)
-if [[ $BUILDKITE_BRANCH == "main" ]]; then
-    file_diff=$(get_diff_main)
-fi
-
-# ----------------------------------------------------------------------
-# Early exit: skip pipeline if all changed files are under docs/
-# ----------------------------------------------------------------------
-if [[ "${DOCS_ONLY_DISABLE}" != "1" ]] && [[ -n "${file_diff:-}" ]]; then
-    docs_only=1
-    while IFS= read -r f; do
-        [[ -z "$f" ]] && continue
-        if [[ "$f" != docs/* ]]; then
-            docs_only=0
-            break
-        fi
-    done < <(printf '%s\n' "$file_diff" | tr ' ' '\n' | tr -d '\r')
-
-    if [[ "$docs_only" -eq 1 ]]; then
-        buildkite-agent annotate ":memo: CI skipped — docs only" --style "info"
-        exit 0
-    fi
-fi
-
-upload_pipeline

.buildkite/common/ci_mirror_hardwares.yml

@@ -0,0 +1,239 @@
+# Uploader-only hardware presets for CUDA / NPU test YAML.
+# Referenced as ``mirror_hardwares: l4_1`` on a step; ``upload_pipeline.py`` expands to
+# ``agents`` (+ optional top-level ``image`` for NPU) + ``plugins``, then strips
+# ``mirror_hardwares`` before Buildkite upload.
+#
+# CUDA naming: {chip}_{count}
+#   l4_1, l4_4, h100_1 … h100_4
+# NPU naming: {queue}_{resource_class}
+#   a2b3_npu_1  — ascend-a2b3, resource_class npu-1 (A2/B3 CI image)
+#   a2b3_npu_4  — ascend-a2b3, resource_class npu-4
+#   a3_npu_2    — ascend-a3, resource_class npu-2 (A3 CI image tag …:$BUILDKITE_COMMIT-a3)
+
+x-npu-k8s-plugins: &npu_k8s_plugins
+  - kubernetes:
+      podSpecPatch:
+        imagePullSecrets:
+          - name: swr-secret
+
+
+mirror_hardwares:
+  l4_1:
+    agents:
+      queue: gpu_1_queue
+    plugins:
+      - docker#v5.2.0:
+          image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
+          always-pull: true
+          propagate-environment: true
+          shm-size: "8gb"
+          environment:
+            - HF_HOME=/fsx/hf_cache
+            - HF_TOKEN
+          volumes:
+            - /fsx/hf_cache:/fsx/hf_cache
+
+  l4_4:
+    agents:
+      queue: gpu_4_queue
+    plugins:
+      - docker#v5.2.0:
+          image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
+          always-pull: true
+          propagate-environment: true
+          shm-size: "8gb"
+          environment:
+            - HF_HOME=/fsx/hf_cache
+            - HF_TOKEN
+          volumes:
+            - /fsx/hf_cache:/fsx/hf_cache
+
+  h100_1:
+    agents:
+      queue: mithril-h100-pool
+    plugins:
+      - kubernetes:
+          podSpec:
+            containers:
+              - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
+                resources:
+                  limits:
+                    nvidia.com/gpu: 1
+                volumeMounts:
+                  - name: devshm
+                    mountPath: /dev/shm
+                  - name: hf-cache
+                    mountPath: /root/.cache/huggingface
+                env:
+                  - name: HF_HOME
+                    value: /root/.cache/huggingface
+                  - name: HF_TOKEN
+                    valueFrom:
+                      secretKeyRef:
+                        name: hf-token-secret
+                        key: token
+            nodeSelector:
+              node.kubernetes.io/instance-type: gpu-h100-sxm
+            volumes:
+              - name: devshm
+                emptyDir:
+                  medium: Memory
+              - name: hf-cache
+                hostPath:
+                  path: /mnt/hf-cache
+                  type: DirectoryOrCreate
+
+  h100_2:
+    agents:
+      queue: mithril-h100-pool
+    plugins:
+      - kubernetes:
+          podSpec:
+            containers:
+              - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
+                resources:
+                  limits:
+                    nvidia.com/gpu: 2
+                volumeMounts:
+                  - name: devshm
+                    mountPath: /dev/shm
+                  - name: hf-cache
+                    mountPath: /root/.cache/huggingface
+                env:
+                  - name: HF_HOME
+                    value: /root/.cache/huggingface
+                  - name: HF_TOKEN
+                    valueFrom:
+                      secretKeyRef:
+                        name: hf-token-secret
+                        key: token
+            nodeSelector:
+              node.kubernetes.io/instance-type: gpu-h100-sxm
+            volumes:
+              - name: devshm
+                emptyDir:
+                  medium: Memory
+              - name: hf-cache
+                hostPath:
+                  path: /mnt/hf-cache
+                  type: DirectoryOrCreate
+
+  h100_3:
+    agents:
+      queue: mithril-h100-pool
+    plugins:
+      - kubernetes:
+          podSpec:
+            containers:
+              - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
+                resources:
+                  limits:
+                    nvidia.com/gpu: 3
+                volumeMounts:
+                  - name: devshm
+                    mountPath: /dev/shm
+                  - name: hf-cache
+                    mountPath: /root/.cache/huggingface
+                env:
+                  - name: HF_HOME
+                    value: /root/.cache/huggingface
+                  - name: HF_TOKEN
+                    valueFrom:
+                      secretKeyRef:
+                        name: hf-token-secret
+                        key: token
+            nodeSelector:
+              node.kubernetes.io/instance-type: gpu-h100-sxm
+            volumes:
+              - name: devshm
+                emptyDir:
+                  medium: Memory
+              - name: hf-cache
+                hostPath:
+                  path: /mnt/hf-cache
+                  type: DirectoryOrCreate
+
+  h100_4:
+    agents:
+      queue: mithril-h100-pool
+    plugins:
+      - kubernetes:
+          podSpec:
+            containers:
+              - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
+                resources:
+                  limits:
+                    nvidia.com/gpu: 4
+                volumeMounts:
+                  - name: devshm
+                    mountPath: /dev/shm
+                  - name: hf-cache
+                    mountPath: /root/.cache/huggingface
+                env:
+                  - name: HF_HOME
+                    value: /root/.cache/huggingface
+                  - name: HF_TOKEN
+                    valueFrom:
+                      secretKeyRef:
+                        name: hf-token-secret
+                        key: token
+            nodeSelector:
+              node.kubernetes.io/instance-type: gpu-h100-sxm
+            volumes:
+              - name: devshm
+                emptyDir:
+                  medium: Memory
+              - name: hf-cache
+                hostPath:
+                  path: /mnt/hf-cache
+                  type: DirectoryOrCreate
+
+  a2b3_npu_1:
+    agents:
+      queue: "ascend-a2b3"
+      resource_class: "npu-1"
+    image: "swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/vllm-omni-ci-npu:${BUILDKITE_COMMIT}"
+    plugins: *npu_k8s_plugins
+
+  a2b3_npu_4:
+    agents:
+      queue: "ascend-a2b3"
+      resource_class: "npu-4"
+    image: "swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/vllm-omni-ci-npu:${BUILDKITE_COMMIT}"
+    plugins: *npu_k8s_plugins
+
+
+  a2b3_npu_8:
+    agents:
+      queue: "ascend-a2b3"
+      resource_class: "npu-8"
+    image: "swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/vllm-omni-ci-npu:${BUILDKITE_COMMIT}"
+    plugins: *npu_k8s_plugins
+
+  a3_npu_2:
+    agents:
+      queue: "ascend-a3"
+      resource_class: "npu-2"
+    image: "swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/vllm-omni-ci-npu:${BUILDKITE_COMMIT}-a3"
+    plugins: *npu_k8s_plugins
+
+  a3_npu_4:
+    agents:
+      queue: "ascend-a3"
+      resource_class: "npu-4"
+    image: "swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/vllm-omni-ci-npu:${BUILDKITE_COMMIT}-a3"
+    plugins: *npu_k8s_plugins
+
+  a3_npu_8:
+    agents:
+      queue: "ascend-a3"
+      resource_class: "npu-8"
+    image: "swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/vllm-omni-ci-npu:${BUILDKITE_COMMIT}-a3"
+    plugins: *npu_k8s_plugins
+
+  a3_npu_16:
+    agents:
+      queue: "ascend-a3"
+      resource_class: "npu-16"
+    image: "swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/vllm-omni-ci-npu:${BUILDKITE_COMMIT}-a3"
+    plugins: *npu_k8s_plugins

.buildkite/common/scripts/resolve_skip_ci.sh

@@ -0,0 +1,23 @@
+#!/usr/bin/env bash
+# Shared Buildkite skip-ci helpers for AMD / Intel bootstrap scripts.
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+SKIP_CI_PY="${SCRIPT_DIR}/skip_ci.py"
+
+# Resolve git diff once. Exit 0 when this bootstrap should stop uploading.
+# Decision details are logged by skip_ci.py; this helper only annotates skip-all.
+# Usage: gate_bootstrap_ci <platform> <l2|l3>
+gate_bootstrap_ci() {
+    local platform="$1"
+    local level="$2"
+    local kind
+    if kind="$(python3 "${SKIP_CI_PY}" gate "${platform}" "${level}")"; then
+        if [[ "${kind}" == "skip-all" ]]; then
+            buildkite-agent annotate \
+                ":memo: CI skipped — docs or pytest skip-mark changes only" \
+                --style "info" 2>/dev/null || true
+        fi
+        exit 0
+    fi
+}

.buildkite/common/scripts/skip_ci.py

@@ -0,0 +1,617 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: Apache-2.0
+"""Shared Buildkite skip-ci and CI-YAML level targeting logic.
+
+``resolve_ci_decision(changed_files, ...)`` buckets explicit file lists (tests, dry-run).
+``resolve_ci_context_from_git()`` resolves git diff once and returns decision + changed files
+for Buildkite bootstrap / upload scripts.
+
+CLI (for bootstrap scripts; exit code is the signal):
+  python3 skip_ci.py gate <platform> <level>
+      Exit 0 when bootstrap should stop (skip-all, or that L2/L3 target is off).
+      Prints ``skip-all`` or ``skip-l23`` on stdout; logs the decision once.
+  python3 skip_ci.py check-skip-all
+      Exit 0 when only docs/skip-mark changes → bootstrap skips entire CI upload.
+  python3 skip_ci.py check-skip-l2-l3 <platform> <level>
+      Exit 0 when that L2/L3 pipeline should be skipped (CI config YAML-only diffs).
+  python3 skip_ci.py print-annotate
+      Print decision message for Buildkite annotation.
+"""
+
+from __future__ import annotations
+
+import argparse
+import os
+import re
+import subprocess
+import sys
+from dataclasses import dataclass, field, replace
+from pathlib import Path
+
+LOG = "[skip-ci]"
+ROOT = Path(__file__).resolve().parent.parent.parent.parent
+
+PLATFORMS = ("cuda", "amd", "intel", "npu")
+L23_LEVELS = frozenset({"l2", "l3"})
+
+L2_YAML_FILES: dict[str, str] = {
+    ".buildkite/cuda/test-ready.yml": "cuda",
+    ".buildkite/npu/test-npu-ready.yml": "npu",
+    ".buildkite/amd/test-amd-ready.yml": "amd",
+    ".buildkite/intel/pipeline-intel.yml": "intel",
+}
+
+L3_YAML_FILES: dict[str, str] = {
+    ".buildkite/cuda/test-merge.yml": "cuda",
+    ".buildkite/amd/test-amd-merge.yml": "amd",
+}
+
+L45_YAML_FILES: dict[str, str] = {
+    ".buildkite/cuda/test-nightly.yml": "cuda",
+    ".buildkite/cuda/test-weekly.yml": "cuda",
+    ".buildkite/npu/test-npu-nightly.yml": "npu",
+}
+
+
+def _empty_platform_buckets() -> dict[str, list[str]]:
+    return {platform: [] for platform in PLATFORMS}
+
+
+@dataclass(frozen=True)
+class CiDevice:
+    """Per-platform / per-level skip flags for yaml-gated bootstrap (``skip_l2_l3``)."""
+
+    skip_cuda_l2: bool = False
+    skip_cuda_l3: bool = False
+    skip_amd_l2: bool = False
+    skip_amd_l3: bool = False
+    skip_intel_l2: bool = False
+    skip_intel_l3: bool = False
+    skip_npu_l2: bool = False
+    skip_npu_l3: bool = False
+
+    def is_skip(self, platform: str, level: str) -> bool:
+        return bool(getattr(self, f"skip_{platform}_{level}", True))
+
+
+@dataclass
+class DiffBuckets:
+    """Changed files grouped by doc / skip-mark / CI yaml / other."""
+
+    docs: list[str] = field(default_factory=list)
+    skip_tests: list[str] = field(default_factory=list)
+    other: list[str] = field(default_factory=list)
+    l2: dict[str, list[str]] = field(default_factory=_empty_platform_buckets)
+    l3: dict[str, list[str]] = field(default_factory=_empty_platform_buckets)
+    l45: dict[str, list[str]] = field(default_factory=_empty_platform_buckets)
+
+    @property
+    def has_doc_changes(self) -> bool:
+        return bool(self.docs)
+
+    @property
+    def has_skip_test_changes(self) -> bool:
+        return bool(self.skip_tests)
+
+    @property
+    def has_other_changes(self) -> bool:
+        """Product code or paths outside skip-ci scope."""
+        return bool(self.other)
+
+    @property
+    def has_l23_yaml_changes(self) -> bool:
+        """At least one whitelisted L2/L3 CI yaml file changed."""
+        return any(self.l2[platform] or self.l3[platform] for platform in PLATFORMS)
+
+    @property
+    def has_l45_yaml_changes(self) -> bool:
+        """At least one whitelisted L4/L5 (nightly/weekly) CI yaml file changed."""
+        return any(self.l45[platform] for platform in PLATFORMS)
+
+    @property
+    def has_skip_ci_scope_changes(self) -> bool:
+        """At least one doc / skip-mark / whitelisted CI yaml change."""
+        return (
+            self.has_doc_changes or self.has_skip_test_changes or self.has_l23_yaml_changes or self.has_l45_yaml_changes
+        )
+
+    def l2_platforms(self) -> frozenset[str]:
+        return frozenset(platform for platform in PLATFORMS if self.l2[platform])
+
+    def l3_platforms(self) -> frozenset[str]:
+        return frozenset(platform for platform in PLATFORMS if self.l3[platform])
+
+
+@dataclass(frozen=True)
+class SkipL2L3Basis:
+    skip_all_l2: bool
+    skip_all_l3: bool
+
+    @classmethod
+    def from_buckets(cls, buckets: DiffBuckets) -> SkipL2L3Basis | None:
+        """L4/L5 yaml → skip both L2 and L3 (L2/L3 buckets may rescue).
+
+        Docs / skip-mark changes do not block this path; they are ignored for
+        L2/L3 targeting when CI YAML is also present.
+        """
+        if buckets.has_other_changes:
+            return None
+        if not buckets.has_l45_yaml_changes:
+            return None
+        return cls(
+            skip_all_l2=not buckets.l2_platforms(),
+            skip_all_l3=not buckets.l3_platforms(),
+        )
+
+
+@dataclass(frozen=True)
+class SkipL3Basis:
+    enable_l2_platforms: frozenset[str]
+
+    @classmethod
+    def from_buckets(cls, buckets: DiffBuckets) -> SkipL3Basis | None:
+        """L2 yaml changed → skip L3; keep L2 on touched platforms."""
+        platforms = buckets.l2_platforms()
+        if not platforms:
+            return None
+        return cls(enable_l2_platforms=platforms)
+
+
+@dataclass(frozen=True)
+class SkipL2Basis:
+    enable_l3_platforms: frozenset[str]
+
+    @classmethod
+    def from_buckets(cls, buckets: DiffBuckets) -> SkipL2Basis | None:
+        """L3 yaml changed → skip L2; keep L3 on touched platforms."""
+        platforms = buckets.l3_platforms()
+        if not platforms:
+            return None
+        return cls(enable_l3_platforms=platforms)
+
+
+@dataclass(frozen=True)
+class CiDecision:
+    """Final CI skip decision for one diff."""
+
+    skip_all: bool = False
+    skip_l2_l3: bool = False
+    device: CiDevice = field(default_factory=CiDevice)
+    message: str = ""
+
+    @classmethod
+    def skip_all_basis(cls, buckets: DiffBuckets, *, diff_range: str | None) -> bool:
+        """Docs / pytest skip-mark only → suppress entire default CI."""
+        if buckets.has_other_changes:
+            return False
+        if (buckets.has_l23_yaml_changes or buckets.has_l45_yaml_changes) and (
+            buckets.has_doc_changes or buckets.has_skip_test_changes
+        ):
+            return False
+        if not buckets.has_doc_changes and not buckets.has_skip_test_changes:
+            return False
+        if diff_range is None and not buckets.has_doc_changes:
+            return False
+        return True
+
+    @classmethod
+    def from_yaml_basis(
+        cls,
+        l2_l3_basis: SkipL2L3Basis | None,
+        l3_basis: SkipL3Basis | None,
+        l2_basis: SkipL2Basis | None,
+    ) -> CiDecision:
+        """Build a yaml-gated decision by merging L2/L3 skip bases.
+
+        Bases are applied as a union of enables: L2 yaml keeps L2 on touched
+        platforms, L3 yaml keeps L3 on touched platforms. Cross-platform mixes
+        (e.g. CUDA ready + AMD merge) must preserve both enables. The resulting
+        per-platform / per-level matrix is stored on ``device`` as ``skip_*``.
+        """
+        run = {platform: {"l2": True, "l3": True} for platform in PLATFORMS}
+
+        if l2_l3_basis is not None:
+            if l2_l3_basis.skip_all_l2:
+                for platform in PLATFORMS:
+                    run[platform]["l2"] = False
+            if l2_l3_basis.skip_all_l3:
+                for platform in PLATFORMS:
+                    run[platform]["l3"] = False
+
+        if l3_basis is not None:
+            for platform in PLATFORMS:
+                run[platform]["l3"] = False
+                run[platform]["l2"] = platform in l3_basis.enable_l2_platforms
+
+        if l2_basis is not None:
+            for platform in PLATFORMS:
+                if platform in l2_basis.enable_l3_platforms:
+                    run[platform]["l3"] = True
+                    if l3_basis is None or platform not in l3_basis.enable_l2_platforms:
+                        run[platform]["l2"] = False
+                elif l3_basis is None:
+                    # L3-yaml-only: disable platforms that did not change L3.
+                    run[platform]["l2"] = False
+                    run[platform]["l3"] = False
+                # else: leave L2 enables from SkipL3Basis; keep this platform's L3 off.
+
+        return cls(
+            skip_l2_l3=True,
+            device=cls._device_from_run(run),
+        )
+
+    @staticmethod
+    def _device_from_run(run: dict[str, dict[str, bool]]) -> CiDevice:
+        return CiDevice(
+            **{f"skip_{platform}_{level}": not run[platform][level] for platform in PLATFORMS for level in ("l2", "l3")}
+        )
+
+    def is_run(self, platform: str, level: str) -> bool:
+        if self.skip_all:
+            return False
+        if not self.skip_l2_l3:
+            return True
+        if level not in L23_LEVELS:
+            return True
+        return not self.device.is_skip(platform, level)
+
+
+@dataclass(frozen=True)
+class CiContext:
+    """Skip-ci decision plus git diff inputs used to compute it."""
+
+    decision: CiDecision
+    changed_files: list[str] | None
+    diff_range: str | None
+
+
+def _log(message: str) -> None:
+    print(f"{LOG}: {message}", file=sys.stderr)
+
+
+def _git(*args: str) -> subprocess.CompletedProcess[str]:
+    return subprocess.run(
+        ["git", *args],
+        check=False,
+        capture_output=True,
+        text=True,
+    )
+
+
+# --- Diff helpers ---
+
+
+def _resolve_diff_range() -> str | None:
+    """Return a git diff range for PR or main builds, or None when unavailable."""
+    is_pr = os.environ.get("BUILDKITE_PULL_REQUEST", "false") != "false" and os.environ.get(
+        "BUILDKITE_PULL_REQUEST", ""
+    )
+    commit = os.environ.get("BUILDKITE_COMMIT", "")
+
+    if is_pr:
+        base_branch = os.environ.get("BUILDKITE_PULL_REQUEST_BASE_BRANCH", "main")
+        base_ref = f"origin/{base_branch}"
+        if _git("rev-parse", "--verify", base_ref).returncode != 0:
+            _log(f"origin/{base_branch} not found locally; trying fetch")
+            _git("fetch", "--depth=200", "origin", base_branch)
+        if _git("rev-parse", "--verify", base_ref).returncode != 0:
+            if _git("rev-parse", "--verify", base_branch).returncode == 0:
+                base_ref = base_branch
+            else:
+                _log(f"cannot resolve PR base {base_branch}; using safe defaults")
+                return None
+        return f"{base_ref}...{commit}"
+
+    if os.environ.get("BUILDKITE_BRANCH", "") == "main":
+        if _git("rev-parse", "--verify", f"{commit}^").returncode != 0:
+            _log("main commit has no parent; using safe defaults")
+            return None
+        return f"{commit}^..{commit}"
+
+    _log("not PR/main build; using safe defaults")
+    return None
+
+
+def _changed_files_for_diff_range(diff_range: str | None) -> list[str] | None:
+    """Return changed file paths for *diff_range*, or None when unavailable."""
+    if diff_range is None:
+        return None
+
+    result = _git("diff", "--name-only", diff_range)
+    if result.returncode != 0:
+        _log(f"git diff failed ({diff_range}); using safe defaults")
+        return None
+
+    files = [line for line in result.stdout.splitlines() if line.strip()]
+    _log(f"{len(files)} changed file(s)")
+    return files
+
+
+def resolve_ci_context_from_git() -> CiContext:
+    """Resolve git diff once and return skip-ci decision plus changed files."""
+    diff_range = _resolve_diff_range()
+    changed_files = _changed_files_for_diff_range(diff_range)
+    decision = resolve_ci_decision(changed_files, diff_range=diff_range)
+    return CiContext(decision=decision, changed_files=changed_files, diff_range=diff_range)
+
+
+# --- Doc / skip-mark helpers ---
+
+
+_SKIP_MARK_RE = re.compile(r"pytest\.mark\.skip(?:if)?\b|pytest\.skip\s*\(")
+_PYTESTMARK_SKIP_RE = re.compile(r"pytest\.mark\.skip\b")
+_PYTEST_MARK_ONLY_RE = re.compile(r"pytest\.mark\.\w+")
+
+
+def _diff_only_contains_skip_mark_changes(diff_text: str) -> bool:
+    """Return True when a unified diff only *adds* pytest skip marks.
+
+    Deleting a skip mark re-enables the test and must not suppress CI. Editing an
+    existing skip (``-`` old / ``+`` new) still qualifies when at least one skip
+    mark is added.
+    """
+
+    def paren_balance(line: str) -> int:
+        return line.count("(") - line.count(")")
+
+    def is_blank_or_comment(line: str) -> bool:
+        stripped = line.strip()
+        return not stripped or stripped.startswith("#")
+
+    def is_skip_mark_line(line: str) -> bool:
+        return _SKIP_MARK_RE.search(line.strip()) is not None
+
+    def is_pytestmark_adjacent_line(line: str) -> bool:
+        stripped = line.strip().rstrip(",")
+        if not stripped or stripped in {"[", "]"} or stripped.startswith("#"):
+            return True
+        if stripped.startswith("pytestmark"):
+            return True
+        return _PYTEST_MARK_ONLY_RE.search(stripped) is not None
+
+    pending_depth = 0
+    saw_change = False
+    has_added_skip_mark = False
+    for raw_line in diff_text.splitlines():
+        if raw_line.startswith("@@"):
+            pending_depth = 0
+            continue
+        if not (raw_line.startswith("+") or raw_line.startswith("-")):
+            continue
+        if raw_line.startswith("+++") or raw_line.startswith("---"):
+            continue
+
+        saw_change = True
+        is_add = raw_line.startswith("+")
+        content = raw_line[1:]
+
+        if pending_depth > 0:
+            pending_depth += paren_balance(content)
+            if pending_depth < 0:
+                pending_depth = 0
+            continue
+
+        if is_blank_or_comment(content):
+            continue
+
+        if is_skip_mark_line(content):
+            if is_add:
+                has_added_skip_mark = True
+            pending_depth = max(0, paren_balance(content))
+            continue
+
+        if is_pytestmark_adjacent_line(content):
+            continue
+
+        return False
+
+    return saw_change and has_added_skip_mark
+
+
+def _is_skip_testcase_change(file_path: str, *, diff_range: str) -> bool:
+    """Return True when a test file diff is skip-mark-only (including new-file cases)."""
+    diff_result = _git("diff", diff_range, "--", file_path)
+    if diff_result.returncode != 0:
+        _log(f"skip-mark-only: git diff failed for {file_path}")
+        return False
+
+    diff_text = diff_result.stdout
+    if not diff_text.strip():
+        _log(f"skip-mark-only: empty diff for {file_path}")
+        return False
+
+    if _diff_only_contains_skip_mark_changes(diff_text):
+        return True
+
+    if "new file mode" not in diff_text and "--- /dev/null" not in diff_text:
+        _log(f"skip-mark-only: non-skip changes in {file_path}")
+        return False
+
+    commit = os.environ.get("BUILDKITE_COMMIT", "")
+    if not commit:
+        return False
+
+    content_result = _git("show", f"{commit}:{file_path}")
+    if content_result.returncode != 0:
+        _log(f"skip-mark-only: cannot read {file_path} at {commit}")
+        return False
+
+    content = content_result.stdout
+    match = re.search(r"^pytestmark\s*=\s*\[(.*?)^\s*\]", content, re.MULTILINE | re.DOTALL)
+    if match is not None and _PYTESTMARK_SKIP_RE.search(match.group(1)) is not None:
+        _log(f"skip-mark-only: new test file with module-level pytest.mark.skip: {file_path}")
+        return True
+    if re.search(r"^pytestmark\s*=\s*pytest\.mark\.skip\b", content, re.MULTILINE):
+        _log(f"skip-mark-only: new test file with module-level pytest.mark.skip: {file_path}")
+        return True
+
+    _log(f"skip-mark-only: new test file without module-level pytest.mark.skip: {file_path}")
+    return False
+
+
+def _classify_changed_files_into_buckets(changed_files: list[str], *, diff_range: str | None) -> DiffBuckets:
+    """Classify each changed file path into a ``DiffBuckets`` category."""
+    buckets = DiffBuckets()
+    for file_path in changed_files:
+        if not file_path:
+            continue
+        if file_path.startswith("docs/") or file_path.endswith(".md") or file_path == "mkdocs.yml":
+            buckets.docs.append(file_path)
+            continue
+        path = Path(file_path)
+        if path.suffix == ".py" and path.parts and path.parts[0] == "tests":
+            if diff_range is not None and _is_skip_testcase_change(file_path, diff_range=diff_range):
+                buckets.skip_tests.append(file_path)
+            else:
+                buckets.other.append(file_path)
+            continue
+        if file_path in L2_YAML_FILES:
+            platform = L2_YAML_FILES[file_path]
+            buckets.l2[platform].append(file_path)
+        elif file_path in L3_YAML_FILES:
+            platform = L3_YAML_FILES[file_path]
+            buckets.l3[platform].append(file_path)
+        elif file_path in L45_YAML_FILES:
+            platform = L45_YAML_FILES[file_path]
+            buckets.l45[platform].append(file_path)
+        else:
+            buckets.other.append(file_path)
+    return buckets
+
+
+# --- Decision ---
+
+
+def _finish(decision: CiDecision, message: str) -> CiDecision:
+    finished = replace(decision, message=message)
+    _log(message)
+    return finished
+
+
+def _format_yaml_gated_message(buckets: DiffBuckets, decision: CiDecision) -> str:
+    """Human-readable summary of which CI YAML changed and which L2/L3 targets run."""
+    changed_parts: list[str] = []
+    for level_name, by_platform in (("L2", buckets.l2), ("L3", buckets.l3), ("L4/L5", buckets.l45)):
+        files = [path for platform in PLATFORMS for path in by_platform[platform]]
+        if files:
+            changed_parts.append(f"{level_name}=[{', '.join(files)}]")
+
+    run_targets = [
+        f"{platform}/{level}" for platform in PLATFORMS for level in ("l2", "l3") if decision.is_run(platform, level)
+    ]
+    skip_targets = [
+        f"{platform}/{level}"
+        for platform in PLATFORMS
+        for level in ("l2", "l3")
+        if not decision.is_run(platform, level)
+    ]
+
+    changed = "; ".join(changed_parts) if changed_parts else "none"
+    run = ", ".join(run_targets) if run_targets else "none"
+    skip = ", ".join(skip_targets) if skip_targets else "none"
+    if buckets.has_doc_changes or buckets.has_skip_test_changes:
+        return (
+            "CI config YAML + docs/skip-mark changed — follow CI YAML gating; "
+            f"changed: {changed}; run: {run}; skip: {skip}"
+        )
+    return f"only CI config YAML changed — changed: {changed}; run: {run}; skip: {skip}"
+
+
+def resolve_ci_decision(
+    changed_files: list[str] | None = None,
+    *,
+    diff_range: str | None = None,
+) -> CiDecision:
+    if changed_files is None:
+        return _finish(CiDecision(), "could not resolve changed files; run normal CI")
+
+    buckets = _classify_changed_files_into_buckets(changed_files, diff_range=diff_range)
+    if buckets.has_other_changes or not buckets.has_skip_ci_scope_changes:
+        return _finish(CiDecision(), "product or non-whitelisted changes present; run normal CI")
+
+    if CiDecision.skip_all_basis(buckets, diff_range=diff_range):
+        return _finish(
+            CiDecision(skip_all=True),
+            "CI skipped — docs or pytest skip-mark changes only",
+        )
+
+    # Docs / skip-mark mixed with whitelisted CI YAML → same yaml-gated path
+    # as CI-YAML-only (docs/skip-mark do not widen to normal CI).
+    if not buckets.has_l23_yaml_changes and not buckets.has_l45_yaml_changes:
+        if buckets.has_skip_test_changes and diff_range is None:
+            message = (
+                "only pytest skip-mark test changes detected, but no git diff range "
+                "is available to verify them (skip-all requires diff_range); run normal CI"
+            )
+        elif buckets.has_skip_test_changes:
+            message = "pytest skip-mark test changes did not qualify for skip-all and no ci yaml changed; run normal CI"
+        else:
+            message = "doc/skip-mark changes did not qualify for skip-all; run normal CI"
+        return _finish(CiDecision(), message)
+
+    decision = CiDecision.from_yaml_basis(
+        SkipL2L3Basis.from_buckets(buckets),
+        SkipL3Basis.from_buckets(buckets),
+        SkipL2Basis.from_buckets(buckets),
+    )
+    return _finish(decision, _format_yaml_gated_message(buckets, decision))
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser(
+        description=__doc__,
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+    subparsers = parser.add_subparsers(dest="command", required=True)
+
+    gate_parser = subparsers.add_parser(
+        "gate",
+        help="exit 0 when bootstrap should stop (skip-all or PLATFORM/LEVEL disabled); logs once",
+    )
+    gate_parser.add_argument("platform", help="cuda, amd, intel, or npu")
+    gate_parser.add_argument("level", choices=("l2", "l3"), help="test level")
+
+    subparsers.add_parser(
+        "check-skip-all",
+        help="exit 0 when docs/skip-mark-only diff should skip entire CI upload",
+    )
+
+    skip_parser = subparsers.add_parser(
+        "check-skip-l2-l3",
+        help="exit 0 when PLATFORM l2/l3 pipeline should be skipped (CI config YAML-only diffs)",
+    )
+    skip_parser.add_argument("platform", help="cuda, amd, intel, or npu")
+    skip_parser.add_argument("level", choices=("l2", "l3"), help="test level")
+
+    subparsers.add_parser(
+        "print-annotate",
+        help="print human-readable skip reason for Buildkite annotation",
+    )
+
+    args = parser.parse_args()
+    decision = resolve_ci_context_from_git().decision
+
+    if args.command == "gate":
+        if decision.skip_all:
+            print("skip-all")
+            return 0
+        if not decision.is_run(args.platform, args.level):
+            print("skip-l23")
+            return 0
+        return 1
+
+    if args.command == "check-skip-all":
+        return 0 if decision.skip_all else 1
+
+    if args.command == "check-skip-l2-l3":
+        return 0 if not decision.is_run(args.platform, args.level) else 1
+
+    if args.command == "print-annotate":
+        print(decision.message or "CI will run normally")
+        return 0
+
+    return 1
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

.buildkite/common/scripts/upload_pipeline.py

@@ -0,0 +1,449 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: Apache-2.0
+"""Render and optionally upload Buildkite pipeline YAML with diff-aware logic.
+
+Bootstrap mode (``bootstrap-upload-steps.yml``):
+  - Hook uploads the entry YAML (``pipeline.yml`` / ``pipeline-npu.yml``) with one step that runs
+    ``upload_pipeline.py --upload <platform>/bootstrap-upload-steps.yml``.
+  - Injects ``if`` by step ``key`` from skip-ci and uploads child steps (image build, L2–L5 upload).
+  - Detect docs-only, pytest skip-mark-only, or combined skip-ci from git diff.
+  - When only CI level YAML changes, enable **L2/L3** upload steps for affected levels only.
+
+Test pipeline mode (e.g. test-merge.yml):
+  - Drop steps whose ``source_file_dependencies`` do not match changed files.
+  - Expand uploader-only ``mirror_hardwares: l4_1`` into ``agents`` (+ optional ``image``
+    for NPU) + ``plugins`` (see ci_mirror_hardwares.yml).
+
+Usage:
+  python3 upload_pipeline.py [--upload] [--all | --e2e] <pipeline.yml>
+
+Requires PyYAML (``pip install pyyaml``); installs it automatically when missing.
+"""
+
+from __future__ import annotations
+
+import argparse
+import copy
+import subprocess
+import sys
+from functools import lru_cache
+from pathlib import Path
+from typing import Any
+
+try:
+    import yaml
+except ModuleNotFoundError:
+    subprocess.run(
+        [sys.executable, "-m", "pip", "install", "-q", "pyyaml"],
+        check=True,
+    )
+    import yaml
+
+from skip_ci import (
+    ROOT,
+    resolve_ci_context_from_git,
+)
+
+# --- Constants ---
+
+LOG = "upload_pipeline"
+BOOTSTRAP_STEPS_FILENAME = "bootstrap-upload-steps.yml"
+BOOTSTRAP_IMAGE_BUILD_KEYS = frozenset({"image-build", "image-build-a2", "image-build-a3"})
+BOOTSTRAP_UPLOAD_IF_KEYS = {
+    "upload-ready-pipeline": "ready",
+    "upload-merge-pipeline": "merge",
+    "upload-nightly-pipeline": "nightly",
+    "upload-weekly-pipeline": "weekly",
+}
+E2E_GROUP_MARKER = "E2E Test"
+CI_MIRROR_HARDWARES_PATH = ROOT / ".buildkite/common/ci_mirror_hardwares.yml"
+
+CUDA_NIGHTLY_ONLY = (
+    '(build.pull_request.labels includes "nightly-test") || (build.branch == "main" && build.env("NIGHTLY") == "1")'
+)
+NPU_NIGHTLY_ONLY = (
+    '(build.branch == "main" && build.env("NIGHTLY") == "1") || '
+    '(build.branch != "main" && ('
+    'build.pull_request.labels includes "nightly-test" || '
+    'build.pull_request.labels includes "omni-test" || '
+    'build.pull_request.labels includes "tts-test" || '
+    'build.pull_request.labels includes "diffusion-x2iat-test" || '
+    'build.pull_request.labels includes "diffusion-x2v-test"'
+    "))"
+)
+
+
+# --- Logging ---
+
+
+def _log(message: str) -> None:
+    print(f"{LOG}: {message}", file=sys.stderr)
+
+
+# --- Bootstrap pipeline (bootstrap-upload-steps.yml) ---
+
+
+def _get_bootstrap_platform(path: Path) -> str:
+    parts = path.as_posix().split("/")
+    return "npu" if "npu" in parts else "cuda"
+
+
+def _load_bootstrap_steps(path: Path) -> str:
+    if path.name != BOOTSTRAP_STEPS_FILENAME:
+        raise ValueError(f"expected {BOOTSTRAP_STEPS_FILENAME}, got {path.name}")
+    return path.read_text(encoding="utf-8")
+
+
+def _format_bootstrap_if(expr: str) -> str:
+    """Return a Buildkite ``if`` string. Buildkite rejects YAML bool ``if`` values."""
+    if expr in ("true", "false"):
+        return expr
+    return f"({expr})"
+
+
+def _compute_bootstrap_if_exprs(*, decision, platform: str) -> dict[str, str]:
+    disabled = "false"
+    nightly_main = 'build.branch == "main" && build.env("NIGHTLY") == "1"'
+    nightly_only = NPU_NIGHTLY_ONLY if platform == "npu" else CUDA_NIGHTLY_ONLY
+
+    if platform == "npu":
+        ready_pr = (
+            'build.branch != "main" && ('
+            'build.pull_request.labels includes "npu-test" || '
+            'build.pull_request.labels includes "ready"'
+            ")"
+        )
+        nightly_label_if = nightly_only
+        weekly_label_if = disabled
+        merge_base = disabled
+    else:
+        ready_pr = 'build.branch != "main" && build.pull_request.labels includes "ready"'
+        merge_main = 'build.branch == "main" && build.env("NIGHTLY") != "1" && build.env("WEEKLY") != "1"'
+        merge_pr = 'build.branch != "main" && build.pull_request.labels includes "merge-test"'
+        nightly_label_if = (
+            '(build.branch == "main" && build.env("NIGHTLY") == "1") || '
+            '(build.branch != "main" && ('
+            'build.pull_request.labels includes "nightly-test" || '
+            'build.pull_request.labels includes "omni-test" || '
+            'build.pull_request.labels includes "tts-test" || '
+            'build.pull_request.labels includes "diffusion-x2iat-test" || '
+            'build.pull_request.labels includes "diffusion-x2v-test"'
+            "))"
+        )
+        weekly_label_if = (
+            '(build.branch == "main" && build.env("WEEKLY") == "1") || '
+            '(build.branch != "main" && build.pull_request.labels includes "weekly-test")'
+        )
+        merge_base = f"({nightly_main}) || (({merge_main}) || ({merge_pr}))"
+
+    ready_base = f"({nightly_main}) || ({ready_pr})"
+    weekly_main = 'build.branch == "main" && build.env("WEEKLY") == "1"'
+
+    if decision.skip_all:
+        # Docs / skip-mark only: no PR-label escape hatch. Main scheduled
+        # NIGHTLY=1 still runs L4 + L2/L3 (--e2e); WEEKLY=1 still runs L5.
+        image_expr = f"({nightly_main}) || ({weekly_main})" if platform == "cuda" else nightly_main
+        ready_expr = nightly_main
+        merge_expr = nightly_main if platform == "cuda" else disabled
+        nightly_expr = nightly_main
+        weekly_expr = weekly_main if platform == "cuda" else disabled
+    elif decision.skip_l2_l3:
+        l2_enabled = decision.is_run("npu", "l2") if platform == "npu" else decision.is_run("cuda", "l2")
+        l3_enabled = platform == "cuda" and decision.is_run("cuda", "l3")
+
+        ready_expr = ready_base if l2_enabled else disabled
+        merge_expr = merge_base if l3_enabled else disabled
+        nightly_expr = nightly_label_if
+        weekly_expr = weekly_label_if if platform == "cuda" else disabled
+
+        image_parts = [f"({nightly_label_if})"]
+        if platform == "cuda":
+            image_parts.append(f"({weekly_label_if})")
+        if l2_enabled:
+            image_parts.insert(0, f"({ready_base})")
+        if l3_enabled:
+            image_parts.insert(1 if l2_enabled else 0, f"({merge_base})")
+        image_expr = " || ".join(image_parts)
+    else:
+        image_expr = "true"
+        ready_expr = ready_base
+        merge_expr = merge_base if platform == "cuda" else disabled
+        nightly_expr = nightly_label_if
+        weekly_expr = weekly_label_if if platform == "cuda" else disabled
+
+    return {
+        "image": _format_bootstrap_if(image_expr),
+        "ready": _format_bootstrap_if(ready_expr),
+        "merge": _format_bootstrap_if(merge_expr),
+        "nightly": _format_bootstrap_if(nightly_expr),
+        "weekly": _format_bootstrap_if(weekly_expr),
+    }
+
+
+def _apply_bootstrap_if(steps: list[Any], if_exprs: dict[str, str]) -> list[Any]:
+    """Inject ``if`` by step key; drop steps that are unconditionally disabled."""
+    kept: list[Any] = []
+    for step in steps:
+        if not isinstance(step, dict):
+            kept.append(step)
+            continue
+        nested = step.get("steps")
+        if isinstance(nested, list):
+            nested_kept = _apply_bootstrap_if(nested, if_exprs)
+            if nested_kept:
+                kept.append({**step, "steps": nested_kept})
+            continue
+        key = step.get("key")
+        if key in BOOTSTRAP_IMAGE_BUILD_KEYS:
+            step["if"] = if_exprs["image"]
+        elif key in BOOTSTRAP_UPLOAD_IF_KEYS:
+            step["if"] = if_exprs[BOOTSTRAP_UPLOAD_IF_KEYS[key]]
+        if_expr = step.get("if")
+        if if_expr == "false":
+            _log(f"omit disabled bootstrap step {key!r}")
+            continue
+        if if_expr == "true":
+            # Unconditional step: omit ``if`` (Buildkite requires string ``if``, not YAML bool).
+            step.pop("if", None)
+        kept.append(step)
+    return kept
+
+
+def _render_bootstrap_pipeline(
+    steps_yaml: str,
+    *,
+    decision,
+    path: Path,
+) -> str:
+    """Load bootstrap steps YAML and inject ``if`` expressions from skip-ci decision."""
+    doc = yaml.safe_load(steps_yaml)
+    if not isinstance(doc, dict):
+        raise ValueError(f"invalid bootstrap steps YAML for {path}")
+    steps = doc.get("steps")
+    if not isinstance(steps, list):
+        raise ValueError(f"bootstrap steps YAML must contain steps: list in {path}")
+
+    platform = _get_bootstrap_platform(path)
+    if_exprs = _compute_bootstrap_if_exprs(decision=decision, platform=platform)
+    doc["steps"] = _apply_bootstrap_if(steps, if_exprs)
+    return yaml.safe_dump(doc, sort_keys=False)
+
+
+# --- Test pipeline (test-ready.yml, test-merge.yml) ---
+
+
+@lru_cache(maxsize=1)
+def _load_mirror_hardwares() -> dict[str, dict[str, Any]]:
+    if not CI_MIRROR_HARDWARES_PATH.is_file():
+        raise FileNotFoundError(f"missing CI mirror_hardwares registry: {CI_MIRROR_HARDWARES_PATH}")
+    doc = yaml.safe_load(CI_MIRROR_HARDWARES_PATH.read_text(encoding="utf-8"))
+    if not isinstance(doc, dict):
+        raise ValueError(f"invalid CI mirror_hardwares registry: {CI_MIRROR_HARDWARES_PATH}")
+    presets = doc.get("mirror_hardwares")
+    if not isinstance(presets, dict):
+        raise ValueError(f"mirror_hardwares must be a mapping in {CI_MIRROR_HARDWARES_PATH}")
+    return presets
+
+
+def _expand_mirror_hardwares(step: dict[str, Any]) -> dict[str, Any]:
+    """Replace uploader-only ``mirror_hardwares`` with preset fields from ci_mirror_hardwares.yml."""
+    hardware = step.get("mirror_hardwares")
+    if hardware is None:
+        return step
+
+    if not isinstance(hardware, str) or not hardware.strip():
+        raise ValueError(
+            f"mirror_hardwares must be a non-empty string in step {_get_step_label(step)!r}",
+        )
+
+    preset = _load_mirror_hardwares().get(hardware)
+    if preset is None:
+        known = ", ".join(sorted(_load_mirror_hardwares()))
+        raise ValueError(
+            f"unknown mirror_hardwares {hardware!r} in step {_get_step_label(step)!r}; known: {known}",
+        )
+
+    if step.get("agents") is not None or step.get("plugins") is not None or step.get("image") is not None:
+        raise ValueError(
+            f"step {_get_step_label(step)!r} sets mirror_hardwares together with agents/plugins/image; "
+            "use mirror_hardwares only",
+        )
+
+    expanded = copy.deepcopy(preset)
+    return {key: value for key, value in step.items() if key != "mirror_hardwares"} | expanded
+
+
+def _match_source_file(changed_files: list[str], prefixes: list[str]) -> bool:
+    for path in changed_files:
+        for prefix in prefixes:
+            normalized = prefix.rstrip("/")
+            if path == normalized or path.startswith(f"{normalized}/"):
+                return True
+    return False
+
+
+def _get_step_label(step: dict[str, Any]) -> str:
+    return str(step.get("group") or step.get("label") or "<step>")
+
+
+def _process_test_steps(
+    steps: list[Any],
+    changed_files: list[str] | None,
+) -> list[Any]:
+    """Drop steps by ``source_file_dependencies`` when *changed_files* is set; always strip that field."""
+    processed: list[Any] = []
+    for step in steps:
+        if not isinstance(step, dict):
+            processed.append(step)
+            continue
+
+        deps = step.get("source_file_dependencies")
+        if deps is not None and not isinstance(deps, list):
+            raise ValueError(
+                f"source_file_dependencies must be a list in step {_get_step_label(step)!r}",
+            )
+        if changed_files is not None and deps is not None and not _match_source_file(changed_files, deps):
+            _log(f"skip {_get_step_label(step)!r} (no changes under {deps})")
+            continue
+
+        nested = step.get("steps")
+        if nested is not None:
+            kept_nested = _process_test_steps(nested, changed_files)
+            if changed_files is not None and not kept_nested:
+                _log(f"omit empty group {_get_step_label(step)!r}")
+                continue
+            new_step = {key: value for key, value in step.items() if key != "source_file_dependencies"}
+            new_step["steps"] = kept_nested
+            processed.append(new_step)
+            continue
+
+        if deps is not None:
+            processed.append(
+                _expand_mirror_hardwares(
+                    {key: value for key, value in step.items() if key != "source_file_dependencies"},
+                ),
+            )
+        else:
+            processed.append(_expand_mirror_hardwares(step))
+
+    return processed
+
+
+def _select_e2e_group_steps(steps: list[Any]) -> list[Any]:
+    """Keep only top-level groups whose name contains ``E2E_GROUP_MARKER``."""
+    selected = [
+        step
+        for step in steps
+        if isinstance(step, dict) and isinstance(step.get("group"), str) and E2E_GROUP_MARKER in step["group"]
+    ]
+    if not selected:
+        _log(f"no group matching {E2E_GROUP_MARKER!r} found")
+    else:
+        _log(f"keep {len(selected)} group(s) matching {E2E_GROUP_MARKER!r}")
+    return selected
+
+
+def _render_test_pipeline(
+    doc: dict[str, Any],
+    changed_files: list[str] | None,
+    *,
+    e2e_only: bool = False,
+) -> dict[str, Any]:
+    """Filter steps by PR diff and strip uploader-only ``source_file_dependencies`` metadata."""
+    steps = doc.get("steps")
+    if not isinstance(steps, list):
+        return doc
+    if e2e_only:
+        steps = _select_e2e_group_steps(steps)
+    steps = _process_test_steps(steps, changed_files)
+    return {**doc, "steps": steps}
+
+
+# --- Entry (read file → bootstrap or test render → YAML string) ---
+
+
+def _render_pipeline(
+    path: Path,
+    *,
+    force_all: bool = False,
+    e2e_only: bool = False,
+) -> str:
+    if path.name == BOOTSTRAP_STEPS_FILENAME:
+        ctx = resolve_ci_context_from_git()
+        continuation = _load_bootstrap_steps(path)
+        return _render_bootstrap_pipeline(
+            continuation,
+            decision=ctx.decision,
+            path=path,
+        )
+
+    text = path.read_text(encoding="utf-8")
+    ctx = resolve_ci_context_from_git()
+    if force_all or e2e_only:
+        changed_files = None
+    else:
+        changed_files = ctx.changed_files
+
+    doc = yaml.safe_load(text)
+    if not isinstance(doc, dict):
+        raise ValueError(f"invalid pipeline YAML: {path}")
+
+    doc = _render_test_pipeline(doc, changed_files, e2e_only=e2e_only)
+    return yaml.safe_dump(doc, sort_keys=False)
+
+
+def _upload_to_buildkite(content: str) -> None:
+    subprocess.run(
+        ["buildkite-agent", "pipeline", "upload"],
+        input=content,
+        text=True,
+        check=True,
+    )
+
+
+# --- CLI ---
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument(
+        "pipeline",
+        nargs="?",
+        default=".buildkite/cuda/bootstrap-upload-steps.yml",
+        help="Pipeline YAML path (default: .buildkite/cuda/bootstrap-upload-steps.yml)",
+    )
+    parser.add_argument(
+        "--upload",
+        action="store_true",
+        help="Pipe rendered YAML to buildkite-agent pipeline upload",
+    )
+    mode = parser.add_mutually_exclusive_group()
+    mode.add_argument(
+        "--all",
+        action="store_true",
+        help="Keep all steps (disable diff-aware skipping)",
+    )
+    mode.add_argument(
+        "--e2e",
+        action="store_true",
+        help="Keep only the E2E Test group",
+    )
+    args = parser.parse_args()
+
+    path = Path(args.pipeline)
+    if not path.is_absolute():
+        path = ROOT / path
+    if not path.is_file():
+        _log(f"missing pipeline file: {path}")
+        return 1
+
+    rendered = _render_pipeline(path, force_all=args.all, e2e_only=args.e2e)
+    if args.upload:
+        _upload_to_buildkite(rendered)
+    else:
+        sys.stdout.write(rendered)
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

.buildkite/cuda/bootstrap-upload-steps.yml

@@ -0,0 +1,55 @@
+# Uploaded by upload_pipeline.py --upload (not the Buildkite hook).
+# ``if`` is injected by step ``key`` from skip-ci at upload time.
+steps:
+  - label: ":docker: Build image"
+    key: image-build
+    commands:
+      - "bash .buildkite/cuda/scripts/build-ci-image.sh"
+    agents:
+      queue: "cpu_queue_premerge"
+
+  # L2 Test — PR ready label (diff-aware) or main+NIGHTLY=1 (--e2e, full E2E group)
+  - label: "Upload Ready Pipeline"
+    depends_on: image-build
+    key: upload-ready-pipeline
+    commands:
+      - |
+        EXTRA=""
+        if [ "${NIGHTLY:-}" = "1" ] && [ "${BUILDKITE_BRANCH}" = "main" ]; then
+          EXTRA="--e2e"
+        fi
+        python3 .buildkite/common/scripts/upload_pipeline.py --upload $$EXTRA .buildkite/cuda/test-ready.yml
+    agents:
+      queue: "cpu_queue_premerge"
+
+  # L3 Test — PR merge-test label (diff-aware) or main+NIGHTLY=1 (--e2e, full E2E group)
+  - label: "Upload Merge Pipeline"
+    depends_on: image-build
+    key: upload-merge-pipeline
+    commands:
+      - |
+        EXTRA=""
+        if [ "${NIGHTLY:-}" = "1" ] && [ "${BUILDKITE_BRANCH}" = "main" ]; then
+          EXTRA="--e2e"
+        fi
+        python3 .buildkite/common/scripts/upload_pipeline.py --upload $$EXTRA .buildkite/cuda/test-merge.yml
+    agents:
+      queue: "cpu_queue_premerge"
+
+  # L4 Test — main+NIGHTLY=1 (scheduled), or PR with specific label (e.g. add label then Rebuild)
+  - label: "Upload Nightly Pipeline"
+    depends_on: image-build
+    key: upload-nightly-pipeline
+    commands:
+      - python3 .buildkite/common/scripts/upload_pipeline.py --upload .buildkite/cuda/test-nightly.yml
+    agents:
+      queue: "cpu_queue_premerge"
+
+  # L5 Test — main+WEEKLY=1 (scheduled), or PR with weekly-test label
+  - label: "Upload Weekly Pipeline"
+    depends_on: image-build
+    key: upload-weekly-pipeline
+    commands:
+      - python3 .buildkite/common/scripts/upload_pipeline.py --upload .buildkite/cuda/test-weekly.yml
+    agents:
+      queue: "cpu_queue_premerge"

.buildkite/cuda/pipeline.yml

@@ -0,0 +1,9 @@
+# Hook uploads this entry YAML; upload_pipeline.py --upload reads bootstrap-upload-steps.yml
+# and injects ``if`` by step ``key`` from skip-ci.
+steps:
+  - label: ":github: Resolve skip-ci & upload pipeline"
+    key: upload-ci-pipeline
+    commands:
+      - "python3 .buildkite/common/scripts/upload_pipeline.py --upload .buildkite/cuda/bootstrap-upload-steps.yml"
+    agents:
+      queue: "cpu_queue_premerge"

.buildkite/cuda/rebase-pipeline.yml

@@ -0,0 +1,35 @@
+steps:
+  - label: ":docker: Build image"
+    key: image-build
+    commands:
+      - "bash .buildkite/cuda/scripts/build-ci-image.sh"
+    agents:
+      queue: "cpu_queue_premerge"
+
+  # L2 Test
+  - label: "Upload Ready Pipeline"
+    depends_on: image-build
+    key: upload-ready-pipeline
+    commands:
+      - python3 .buildkite/common/scripts/upload_pipeline.py --upload --all .buildkite/cuda/test-ready.yml
+    agents:
+      queue: "cpu_queue_premerge"
+
+  # L3 Test
+  - label: "Upload Merge Pipeline"
+    depends_on: image-build
+    key: upload-merge-pipeline
+    commands:
+      - python3 .buildkite/common/scripts/upload_pipeline.py --upload --all .buildkite/cuda/test-merge.yml
+    agents:
+      queue: "cpu_queue_premerge"
+
+  # L4 Test
+  - label: "Upload Nightly Pipeline"
+    depends_on: image-build
+    key: upload-nightly-pipeline
+    if: build.env("NIGHTLY") == "1"
+    commands:
+      - python3 .buildkite/common/scripts/upload_pipeline.py --upload --all .buildkite/cuda/test-nightly.yml
+    agents:
+      queue: "cpu_queue_premerge"

.buildkite/cuda/scripts/build-ci-image.sh

@@ -0,0 +1,52 @@
+#!/bin/bash
+set -euo pipefail
+
+# Build and push the CI Docker image with registry-based layer caching.
+#
+# Usage: build-ci-image.sh
+#
+# Expects the following environment variables (set by Buildkite):
+#   BUILDKITE_COMMIT          - commit SHA used as the image tag
+ECR_NAMESPACE="public.ecr.aws/q9t5s3a7"
+REGISTRY="${ECR_NAMESPACE}/vllm-ci-test-repo"
+REGION="us-east-1"
+DOCKERFILE="docker/Dockerfile.ci"
+BUILDER_NAME="vllm-omni-builder"
+
+# Ensure that the env vars are actually set, otherwise exit early
+if [ -z "${BUILDKITE_COMMIT:-}" ]; then
+    echo "ERROR: BUILDKITE_COMMIT is not set"
+    exit 1
+fi
+echo "BUILDKITE_COMMIT: ${BUILDKITE_COMMIT}"
+
+# Authenticate to ECR Public
+aws ecr-public get-login-password --region "$REGION" \
+    | docker login --username AWS --password-stdin "$ECR_NAMESPACE"
+
+# Compute the cache tag for the dependencies image as the hash of the contents.
+# This is because dependencies don't change often, so most PRs will have the same hash.
+#
+# The Dockerfile is also included to avoid churn in the cached image based on the vLLM version.
+DEP_FILES="pyproject.toml setup.py requirements/*.txt docker/Dockerfile.ci"
+CACHE_KEY=$(cat ${DEP_FILES} | sha256sum | cut -c1-16)
+CACHE_TAG="deps-cache-${CACHE_KEY}"
+echo "Cache key: ${CACHE_TAG}"
+
+# Set up buildx with docker-container driver; we need to do this
+# since cache export is not supported for the default docker driver
+# that is running in the CI.
+docker buildx inspect $BUILDER_NAME >/dev/null 2>&1 \
+|| docker buildx create --name $BUILDER_NAME --driver docker-container
+docker buildx use $BUILDER_NAME
+
+echo "Building image tag: ${REGISTRY}:${BUILDKITE_COMMIT}"
+
+# Build + push the image to the registry, using the hash from the dependencies.
+# Note that the caching here is layerwise, so we're just making sure we always
+# can cache hit on everything but the last copy at the moment.
+docker buildx build --push --progress=plain \
+    --cache-from "type=registry,ref=${REGISTRY}:${CACHE_TAG}" \
+    --cache-to "type=registry,ref=${REGISTRY}:${CACHE_TAG},mode=max,compression=zstd" \
+    --file "$DOCKERFILE" \
+    -t "${REGISTRY}:${BUILDKITE_COMMIT}" .

.buildkite/cuda/test-merge.yml

@@ -0,0 +1,278 @@
+env:
+  VLLM_WORKER_MULTIPROC_METHOD: spawn
+  VLLM_USE_FLASHINFER_MOE_FP16: 0
+  HF_HUB_DOWNLOAD_TIMEOUT: 300
+  HF_HUB_ETAG_TIMEOUT: 60
+
+steps:
+  - group: ":card_index_dividers: Simple Test"
+    depends_on: upload-merge-pipeline
+    steps:
+      - label: "Simple · Diffusion & Model Executor Test"
+        commands:
+          - "timeout 20m pytest -sv tests/diffusion tests/model_executor -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+        mirror_hardwares: l4_1
+
+      - label: "Simple · Other Test"
+        commands:
+          - "timeout 20m pytest -sv -m 'core_model and cpu' --ignore=tests/diffusion --ignore=tests/model_executor --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+        mirror_hardwares: l4_1
+
+  - group: ":card_index_dividers: Diffusion Test"
+    depends_on: upload-merge-pipeline
+    steps:
+      - label: "Diffusion · LoRA Test"
+        timeout_in_minutes: 30
+        commands:
+          - pytest -sv tests/diffusion/lora/ -m 'advanced_model and cuda' --run-level "advanced_model"
+        mirror_hardwares: l4_1
+
+      - label: "Diffusion · Offloader Test"
+        timeout_in_minutes: 40
+        commands:
+          - timeout 40m pytest -s -v tests/diffusion/offloader -m "advanced_model and cuda"
+        mirror_hardwares: l4_1
+
+  - label: "Distributed Test"
+    depends_on: upload-merge-pipeline
+    timeout_in_minutes: 60
+    commands:
+      - |
+        timeout 55m bash -c "
+          set -e
+          export VLLM_IMAGE_FETCH_TIMEOUT=60
+          pytest -s -v tests/distributed -m 'advanced_model and cuda and H100' --run-level 'advanced_model'
+        "
+    mirror_hardwares: h100_1
+
+  - group: ":card_index_dividers: Entrypoints Test"
+    depends_on: upload-merge-pipeline
+    steps:
+      - label: "Entrypoint Test with H100"
+        # TEMPORARY 60 -> 90: engine startup/sleep-wake is ~13-15 min slower per
+        # run on the vLLM 0.26 image (model loading ~2x, init ~2x; see builds
+        # 2915 vs 2916), which pushed this suite past 60 min. Revert once the
+        # 0.26 startup regression is understood and fixed.
+        timeout_in_minutes: 90
+        commands:
+          - pytest -s -v tests/entrypoints/ -m "advanced_model and cuda and H100" --run-level "advanced_model"
+        mirror_hardwares: h100_2
+
+      - label: "Entrypoint Test with L4"
+        timeout_in_minutes: 60
+        commands:
+          - pytest -s -v tests/entrypoints/ -m "advanced_model and cuda and L4" --run-level "advanced_model"
+        mirror_hardwares: l4_1
+
+  - group: ":card_index_dividers: E2E Test"
+    depends_on: upload-merge-pipeline
+    steps:
+      - label: "TTS · Qwen3-TTS CustomVoice Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen3_tts_customvoice.py
+          - tests/e2e/offline_inference/test_qwen3_tts_customvoice.py
+          - vllm_omni/model_executor/models/qwen3_tts/
+          - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
+          - vllm_omni/deploy/qwen3_tts.yaml
+        commands:
+          - |
+            timeout 40m bash -c "
+              export VLLM_LOGGING_LEVEL=DEBUG
+              export VLLM_ALLOW_LONG_MAX_MODEL_LEN=1
+              pytest -s -v tests/e2e/online_serving/test_qwen3_tts_customvoice.py tests/e2e/offline_inference/test_qwen3_tts_customvoice.py -m 'advanced_model and cuda' --run-level 'advanced_model'
+            "
+        mirror_hardwares: l4_1
+
+      - label: "TTS · Qwen3-TTS Base Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen3_tts_base.py
+          - tests/e2e/offline_inference/test_qwen3_tts_base.py
+          - vllm_omni/model_executor/models/qwen3_tts/
+          - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
+          - vllm_omni/deploy/qwen3_tts.yaml
+        commands:
+          - |
+            timeout 40m bash -c "
+              export VLLM_LOGGING_LEVEL=DEBUG
+              export VLLM_ALLOW_LONG_MAX_MODEL_LEN=1
+              pytest -s -v tests/e2e/online_serving/test_qwen3_tts_base.py tests/e2e/offline_inference/test_qwen3_tts_base.py -m 'advanced_model and cuda' --run-level 'advanced_model'
+            "
+        mirror_hardwares: l4_1
+
+      - label: "TTS · Higgs-Audio-V3 Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_higgs_audio_v3.py
+          - vllm_omni/model_executor/models/higgs_audio_v3/
+          - vllm_omni/model_executor/stage_input_processors/higgs_audio_v3.py
+          - vllm_omni/deploy/higgs_multimodal_qwen3.yaml
+        commands:
+          - |
+            timeout 20m bash -c "
+              export VLLM_WORKER_MULTIPROC_METHOD=spawn
+              export VLLM_USE_DEEP_GEMM=0
+              export VLLM_MOE_USE_DEEP_GEMM=0
+              pytest -s -v tests/e2e/online_serving/test_higgs_audio_v3.py -m 'advanced_model' --run-level 'advanced_model'
+            "
+        mirror_hardwares: l4_1
+
+      - label: "Omni · Qwen3-Omni Test"
+        source_file_dependencies:
+          - tests/e2e/offline_inference/test_qwen3_omni.py
+          - tests/e2e/online_serving/test_qwen3_omni.py
+          - vllm_omni/model_executor/models/qwen3_omni/
+          - vllm_omni/model_executor/stage_input_processors/qwen3_omni.py
+          - vllm_omni/deploy/qwen3_omni_moe.yaml
+        timeout_in_minutes: 50
+        commands:
+          - pytest -s -v tests/e2e/offline_inference/test_qwen3_omni.py tests/e2e/online_serving/test_qwen3_omni.py -m 'advanced_model and cuda' --run-level 'advanced_model'
+        mirror_hardwares: h100_2
+
+      - label: "Omni · MiniCPM-o 4.5 Test"
+        source_file_dependencies:
+          - tests/e2e/offline_inference/test_minicpmo_4_5.py
+          - tests/e2e/online_serving/test_minicpmo_4_5.py
+          - vllm_omni/deploy/minicpmo_4_5.yaml
+        timeout_in_minutes: 50
+        commands:
+          - pytest -s -v tests/e2e/offline_inference/test_minicpmo_4_5.py tests/e2e/online_serving/test_minicpmo_4_5.py -m 'advanced_model and cuda' --run-level 'advanced_model'
+        agents:
+          queue: "mithril-h100-pool"
+        plugins:
+          - kubernetes:
+              podSpec:
+                containers:
+                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
+                    resources:
+                      limits:
+                        nvidia.com/gpu: 2
+                    volumeMounts:
+                      - name: devshm
+                        mountPath: /dev/shm
+                      - name: hf-cache
+                        mountPath: /root/.cache/huggingface
+                    env:
+                      - name: HF_HOME
+                        value: /root/.cache/huggingface
+                      - name: HF_TOKEN
+                        valueFrom:
+                          secretKeyRef:
+                            name: hf-token-secret
+                            key: token
+                nodeSelector:
+                  node.kubernetes.io/instance-type: gpu-h100-sxm
+                volumes:
+                  - name: devshm
+                    emptyDir:
+                      medium: Memory
+                  - name: hf-cache
+                    hostPath:
+                      path: /mnt/hf-cache
+                      type: DirectoryOrCreate
+
+      - label: "TTS · VoxCPM2 Test"
+        source_file_dependencies:
+          - tests/e2e/offline_inference/test_voxcpm2_tts.py
+          - tests/e2e/online_serving/test_voxcpm2_tts.py
+          - vllm_omni/model_executor/models/voxcpm2/
+          - vllm_omni/deploy/voxcpm2.yaml
+        timeout_in_minutes: 30
+        commands:
+          - |
+            timeout 30m bash -c "
+              export VLLM_LOGGING_LEVEL=DEBUG
+              export VLLM_WORKER_MULTIPROC_METHOD=spawn
+              pytest -s -v tests/e2e/offline_inference/test_voxcpm2_tts.py tests/e2e/online_serving/test_voxcpm2_tts.py -m 'advanced_model and cuda' --run-level 'advanced_model'
+            "
+        mirror_hardwares: l4_1
+
+      - label: "TTS · SoulX-Singer Test"
+        source_file_dependencies:
+          - tests/e2e/offline_inference/test_soulxsinger.py
+          - vllm_omni/diffusion/models/soulx_singer/
+          - vllm_omni/deploy/soulxsinger_svc.yaml
+          - vllm_omni/deploy/soulxsinger_svs.yaml
+        timeout_in_minutes: 30
+        commands:
+          - |
+            timeout 30m bash -c "
+              export VLLM_LOGGING_LEVEL=DEBUG
+              export VLLM_WORKER_MULTIPROC_METHOD=spawn
+              pytest -s -v tests/e2e/offline_inference/test_soulxsinger.py -k 'svc or precomputed' -m 'advanced_model and cuda' --run-level 'advanced_model'
+            "
+        mirror_hardwares: l4_1
+
+      - label: "Diffusion · Qwen Image Edit Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen_image_edit.py
+          - vllm_omni/diffusion/models/qwen_image/
+        timeout_in_minutes: 20
+        commands:
+          - pytest -s -v tests/e2e/online_serving/test_qwen_image_edit.py -m 'advanced_model and cuda' --run-level 'advanced_model'
+        mirror_hardwares: h100_1
+      - label: "Diffusion · Bagel Test"
+        source_file_dependencies:
+          - tests/e2e/offline_inference/test_bagel.py
+          - tests/e2e/online_serving/test_bagel.py
+          - vllm_omni/model_executor/models/bagel/
+          - vllm_omni/diffusion/models/bagel/
+          - vllm_omni/model_executor/stage_input_processors/bagel.py
+          - vllm_omni/deploy/bagel.yaml
+        timeout_in_minutes: 20
+        commands:
+          - pytest -s -v tests/e2e/offline_inference/test_bagel.py tests/e2e/online_serving/test_bagel.py -m 'advanced_model and cuda' --run-level 'advanced_model'
+        mirror_hardwares: h100_1
+
+      - label: "Diffusion · Qwen Image Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen_image.py
+          - vllm_omni/diffusion/models/qwen_image/
+        commands:
+          - |
+            timeout 40m bash -c '
+              export VLLM_IMAGE_FETCH_TIMEOUT=60
+              pytest -s -v tests/e2e/online_serving/test_qwen_image.py -m 'advanced_model and cuda' --run-level 'advanced_model'
+            '
+        mirror_hardwares: h100_1
+
+      - label: "Diffusion · Qwen Image Layered Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen_image_layered.py
+          - vllm_omni/diffusion/models/qwen_image/
+        commands:
+          - |
+            timeout 40m bash -c "
+              export VLLM_IMAGE_FETCH_TIMEOUT=60
+              pytest -s -v tests/e2e/online_serving/test_qwen_image_layered.py -m 'advanced_model and cuda' --run-level 'advanced_model'
+            "
+        mirror_hardwares: h100_1
+
+      - label: "Diffusion · Wan22 Test"
+        source_file_dependencies:
+          - tests/e2e/offline_inference/test_wan22_t2v.py
+          - tests/e2e/online_serving/test_wan22_t2v.py
+          - vllm_omni/diffusion/models/wan2_2/
+        commands:
+          - |
+            timeout 40m bash -c "
+              export VLLM_IMAGE_FETCH_TIMEOUT=60
+              pytest -s -v tests/e2e/offline_inference/test_wan22_t2v.py tests/e2e/online_serving/test_wan22_t2v.py -m 'advanced_model and cuda and diffusion' --run-level 'advanced_model'
+            "
+        mirror_hardwares: h100_1
+
+      - label: "Diffusion · Z Image Test"
+        source_file_dependencies:
+          - tests/e2e/offline_inference/test_z_image.py
+          - tests/e2e/online_serving/test_z_image.py
+          - vllm_omni/diffusion/models/z_image/
+        timeout_in_minutes: 30
+        commands:
+          - pytest -s -v tests/e2e/offline_inference/test_z_image.py tests/e2e/online_serving/test_z_image.py -m 'advanced_model and cuda and diffusion' --run-level 'advanced_model'
+        mirror_hardwares: l4_1
+      - label: "Diffusion · FLUX.2 Klein Test"
+        source_file_dependencies:
+          - tests/e2e/offline_inference/test_flux2_klein.py
+          - vllm_omni/diffusion/models/flux2_klein/
+        timeout_in_minutes: 30
+        commands:
+          - pytest -s -v tests/e2e/offline_inference/test_flux2_klein.py -m 'advanced_model and cuda and diffusion' --run-level 'advanced_model'
+        mirror_hardwares: l4_1

.buildkite/cuda/test-nightly.yml

@@ -0,0 +1,534 @@
+env:
+  VLLM_WORKER_MULTIPROC_METHOD: spawn
+  VLLM_USE_FLASHINFER_MOE_FP16: 0
+  HF_HUB_DOWNLOAD_TIMEOUT: 300
+  HF_HUB_ETAG_TIMEOUT: 60
+
+steps:
+  # Group: collapses under one heading in the Buildkite UI; child steps still run in parallel.
+  - group: ":card_index_dividers: Omni Model Test"
+    key: nightly-omni-test-group
+    depends_on: upload-nightly-pipeline
+    if: build.env("NIGHTLY") == "1" || build.pull_request.labels includes "nightly-test" || build.pull_request.labels includes "omni-test"
+    steps:
+
+      - label: ":full_moon: Omni · Function Test with H100"
+        timeout_in_minutes: 120
+        commands:
+          - pytest -sv tests/e2e --run-level "full_model" -m "full_model and H100 and omni" --ignore=tests/e2e/accuracy --ignore=tests/e2e/online_serving/test_qwen3_omni_multi_replicas.py
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Omni · Doc Test with L4"
+        timeout_in_minutes: 90
+        commands:
+          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
+          - pytest -s -v tests/examples/ -m "full_model and omni and L4" --run-level "full_model"
+        mirror_hardwares: l4_4
+
+      - label: ":full_moon: Omni · Doc Test with H100"
+        timeout_in_minutes: 90
+        commands:
+          - pytest -s -v tests/examples/ -m "full_model and omni and H100" --run-level "full_model"
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Omni · Accuracy Test"
+        timeout_in_minutes: 180
+        commands:
+          - export SEED_TTS_WER_EVAL=1
+          - export SEED_TTS_EVAL_DEVICE=cuda:1
+          - |
+            set +e
+            pytest -s -v tests/e2e/accuracy/qwen3_omni/test_qwen3_omni.py -m "full_model" --run-level full_model
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/e2e/accuracy/qwen3_omni/results/qwen_omni_acc/*.json"
+            exit $$EXIT
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Omni · Perf Test · No Async Chunk"
+        key: nightly-omni-performance-no-async-chunk
+        timeout_in_minutes: 300
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen3_omni_no_async_chunk.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Omni · Perf Test · Async Chunk"
+        key: nightly-omni-performance-async-chunk
+        timeout_in_minutes: 180
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen3_omni_async_chunk.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Omni · Perf Test · vLLM Text"
+        key: nightly-omni-performance-vllm-text
+        timeout_in_minutes: 180
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen3_omni_vllm_text.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Omni · Perf Test · Multi-Replica"
+        key: nightly-omni-performance-multi-replicas
+        timeout_in_minutes: 180
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen3_omni_multi_replicas.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+        mirror_hardwares: h100_3
+
+      - label: ":full_moon: Omni · Multi-Replica Startup Test with 4x H100"
+        timeout_in_minutes: 45
+        commands:
+          - pytest -s -v tests/e2e/online_serving/test_qwen3_omni_multi_replicas.py -m "full_model" --run-level "core_model"
+        mirror_hardwares: h100_4
+
+  - group: ":card_index_dividers: TTS Model Test"
+    key: nightly-tts-test-group
+    depends_on: upload-nightly-pipeline
+    if: build.env("NIGHTLY") == "1" || build.pull_request.labels includes "nightly-test" || build.pull_request.labels includes "tts-test"
+    steps:
+
+      - label: ":full_moon: TTS · Function Test with L4"
+        timeout_in_minutes: 120
+        commands:
+          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
+          - export VLLM_USE_DEEP_GEMM="0"
+          - export VLLM_MOE_USE_DEEP_GEMM="0"
+          - pytest -s -v tests/e2e/ -m "full_model and L4 and tts" --run-level "full_model" --ignore=tests/e2e/accuracy --ignore=tests/e2e/online_serving/test_voxcpm2_tts_expansion.py
+        mirror_hardwares: l4_1
+
+      - label: ":full_moon: TTS · Accuracy Test · VoxCPM2"
+        timeout_in_minutes: 30
+        commands:
+          - |
+            timeout 30m bash -c "
+              export VLLM_LOGGING_LEVEL=DEBUG
+              export VLLM_WORKER_MULTIPROC_METHOD=spawn
+              export VLLM_ALLOW_LONG_MAX_MODEL_LEN=1
+              pytest -s -v tests/e2e/online_serving/test_voxcpm2_tts_expansion.py -m 'full_model and cuda and tts' --run-level 'full_model'
+            "
+        mirror_hardwares: l4_1
+
+      - label: ":full_moon: TTS · Perf Test"
+        key: nightly-tts-performance
+        timeout_in_minutes: 180
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_tts.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: TTS · VoxCPM2 · Perf Test"
+        key: nightly-tts-performance-voxcpm2
+        timeout_in_minutes: 180
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_voxcpm2.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+        mirror_hardwares: h100_1
+
+      - label: ":full_moon: TTS · Higgs-Audio-V3 · Perf Test"
+        key: nightly-tts-performance-higgs-audio-v3
+        timeout_in_minutes: 180
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - export VLLM_USE_DEEP_GEMM="0"
+          - export VLLM_MOE_USE_DEEP_GEMM="0"
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_higgs_audio_v3.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+        mirror_hardwares: h100_1
+
+  - group: ":card_index_dividers: Tiny Model Tests [Multi-GPU]"
+    key: nightly-tiny-model-test-group
+    depends_on: upload-nightly-pipeline
+    if: >-
+      build.env("NIGHTLY") == "1" ||
+      build.pull_request.labels includes "nightly-test"
+    steps:
+
+      - label: ":full_moon: Tiny Model Tests · Diffusion (Multi-GPU accelerations)"
+        timeout_in_minutes: 30
+        # NOTE: Single-GPU (core_model) tiny model tests run on the ready label,
+        # so nightly only adds the multi-GPU parallelism configurations.
+        #
+        # This is a bit confusing, but `full_model` is used to mark the multigpu tests, and
+        # --run-level `core_model` is used to indicate that we should use tiny model weights
+        # for the CI, since these tests should pass irrespective of whether or not we use
+        # the full weights or not.
+        commands:
+          - timeout 25m pytest -sv tests/model_tests/diffusion -m 'full_model and cuda' --run-level "core_model"
+        mirror_hardwares: l4_4
+
+  - group: ":card_index_dividers: Diffusion X2I(&A&T) Model Test"
+    key: nightly-diffusion-x2iat-group
+    depends_on: upload-nightly-pipeline
+    if: >-
+      build.env("NIGHTLY") == "1" ||
+      build.pull_request.labels includes "nightly-test" ||
+      build.pull_request.labels includes "diffusion-x2iat-test"
+    steps:
+      # Use explicit file shards so single-card cases do not reserve 2+ GPUs,
+      # and Bagel feature coverage is not duplicated via a broad tests/e2e sweep.
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · Function Test with H100 · Single-GPU"
+        timeout_in_minutes: 120
+        commands:
+          - >-
+            pytest -sv
+            tests/e2e/online_serving/test_qwen_image_expansion.py
+            tests/e2e/online_serving/test_qwen_image_edit_expansion.py
+            tests/e2e/online_serving/test_qwen_image_layered_expansion.py
+            tests/e2e/online_serving/test_boogu_image.py
+            tests/e2e/online_serving/test_boogu_image_edit.py
+            tests/e2e/online_serving/test_boogu_image_expansion.py
+            tests/e2e/online_serving/test_boogu_image_edit_expansion.py
+            tests/e2e/online_serving/test_bagel_expansion.py
+            tests/e2e/online_serving/test_krea2_expansion.py
+            tests/e2e/offline_inference/test_krea2_expansion.py
+            -m "full_model and diffusion and H100 and not distributed_cuda"
+            --run-level "full_model"
+        mirror_hardwares: h100_1
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · Function Test with H100 · Multi-GPU Qwen-Image"
+        timeout_in_minutes: 120
+        commands:
+          - >-
+            pytest -sv
+            tests/e2e/online_serving/test_qwen_image_expansion.py
+            tests/e2e/online_serving/test_qwen_image_edit_expansion.py
+            tests/e2e/offline_inference/test_krea2_expansion.py
+            -m "full_model and diffusion and H100 and distributed_cuda"
+            --run-level "full_model"
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · Function Test with H100 · Multi-GPU Layered/LongCat/FLUX"
+        timeout_in_minutes: 90
+        commands:
+          - >-
+            pytest -sv
+            tests/e2e/online_serving/test_qwen_image_layered_expansion.py
+            -m "full_model and diffusion and H100 and distributed_cuda"
+            --run-level "full_model"
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · Function Test with H100 · Multi-GPU Bagel"
+        timeout_in_minutes: 120
+        commands:
+          - >-
+            pytest -sv
+            tests/e2e/online_serving/test_bagel_expansion.py
+            -m "full_model and diffusion and H100 and distributed_cuda"
+            --run-level "full_model"
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · GLM-Image Function Test with H100"
+        timeout_in_minutes: 120
+        commands:
+          - >-
+            pytest -sv
+            tests/e2e/online_serving/test_glm_image_expansion.py
+            tests/e2e/offline_inference/test_glm_image_autoround_w4a16_expansion.py
+            -m "full_model and diffusion and H100"
+            --run-level "full_model"
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · Function Test with L4"
+        timeout_in_minutes: 120
+        commands:
+          - pytest -sv tests/e2e/ -k "not test_wan and not test_bagel_expansion and not hunyuan" -m "full_model and diffusion and L4" --run-level "full_model" --ignore=tests/e2e/accuracy
+        mirror_hardwares: l4_4
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · Doc Test"
+        timeout_in_minutes: 60
+        commands:
+          - pytest -s -v tests/examples/*/test_text_to_image.py -m "full_model and example and H100" --run-level "full_model"
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · Accuracy Test"
+        timeout_in_minutes: 180
+        commands:
+          - export VLLM_HTTP_TIMEOUT_KEEP_ALIVE=120
+          - pytest -s -v tests/e2e/accuracy/test_qwen_image*.py --run-level full_model
+          - pytest -s -v tests/e2e/accuracy/test_diffusers_backend_similarity.py -k '2i_matches_diffusers' -m full_model --run-level full_model
+        mirror_hardwares: h100_1
+
+      - label: ":full_moon: HunyuanImage3-DIT · Accuracy Test"
+        timeout_in_minutes: 180
+        commands:
+          - export HUNYUAN_IMAGE3_MODEL="tencent/HunyuanImage-3.0-Instruct"
+          - export HUNYUAN_IMAGE3_DEVICES="0,1,2,3"
+          - pytest -s -v tests/e2e/accuracy/test_hunyuan_image3_pixel_accuracy.py -m full_model --run-level full_model
+        mirror_hardwares: h100_4
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · Perf Test · Qwen-Image"
+        key: nightly-diffusion-x2iat-performance-qwen-image
+        timeout_in_minutes: 180
+        commands:
+          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
+          - export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN
+          - export CACHE_DIT_VERSION=1.3.0
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen_image_vllm_omni.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
+            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
+            exit $$EXIT
+        mirror_hardwares: h100_4
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · Perf Test · Qwen-Image-Edit"
+        key: nightly-diffusion-x2iat-performance-qwen-image-edit
+        timeout_in_minutes: 180
+        commands:
+          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
+          - export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN
+          - export CACHE_DIT_VERSION=1.3.0
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen_image_edit_vllm_omni.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
+            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
+            exit $$EXIT
+        mirror_hardwares: h100_4
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · Perf Test · Qwen-Image-Edit-2511"
+        key: nightly-diffusion-x2iat-performance-qwen-image-edit-2511
+        timeout_in_minutes: 180
+        commands:
+          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
+          - export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN
+          - export CACHE_DIT_VERSION=1.3.0
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen_image_edit_2511_vllm_omni.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
+            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
+            exit $$EXIT
+        mirror_hardwares: h100_4
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · Perf Test · Qwen-Image-Layered"
+        key: nightly-diffusion-x2iat-performance-qwen-image-layered
+        timeout_in_minutes: 120
+        commands:
+          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
+          - export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN
+          - export CACHE_DIT_VERSION=1.3.0
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen_image_layered_vllm_omni.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
+            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
+            exit $$EXIT
+        mirror_hardwares: h100_1
+
+      - label: ":full_moon: Diffusion X2I(&A&T) · Perf Test · BAGEL"
+        key: nightly-diffusion-x2iat-performance-bagel
+        timeout_in_minutes: 180
+        commands:
+          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
+          - export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN
+          - export CACHE_DIT_VERSION=1.3.0
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py --test-config-file tests/dfx/perf/tests/test_bagel_vllm_omni.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
+            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
+            exit $$EXIT
+        mirror_hardwares: h100_1
+
+  - group: ":card_index_dividers: Diffusion X2V Model Test"
+    key: nightly-diffusion-x2v-group
+    depends_on: upload-nightly-pipeline
+    if: >-
+      build.env("NIGHTLY") == "1" ||
+      build.pull_request.labels includes "nightly-test" ||
+      build.pull_request.labels includes "diffusion-x2v-test"
+    steps:
+
+      - label: ":full_moon: Diffusion X2V · Wan2.2 T2V Function Test"
+        timeout_in_minutes: 90
+        commands:
+          - pytest -s -v tests/e2e/online_serving/test_wan22_expansion.py -k "t2v" -m "full_model and cuda" --run-level "full_model"
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Diffusion X2V · Wan2.2 I2V/TI2V Function Test"
+        timeout_in_minutes: 90
+        commands:
+          - pytest -s -v tests/e2e/online_serving/test_wan22_expansion.py -k "i2v" -m "full_model and cuda" --run-level "full_model"
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Diffusion X2V · LingBot Video MoE Function Test"
+        timeout_in_minutes: 90
+        commands:
+          - pytest -s -v tests/e2e/online_serving/test_lingbot_video_moe.py -m "full_model and cuda" --run-level "full_model"
+        mirror_hardwares: h100_1
+
+      - label: ":full_moon: Diffusion X2V · Doc Test"
+        timeout_in_minutes: 60
+        commands:
+          - pytest -s -v tests/examples/offline_inference/test_image_to_video.py -m "full_model and example and H100" --run-level "full_model"
+        mirror_hardwares: h100_1
+
+      - label: ":full_moon: Diffusion X2V · Other Function Test"
+        timeout_in_minutes: 90
+        commands:
+          - pytest -s -v tests/e2e/offline_inference/test_wan22_autoround_w4a16_expansion.py -m "full_model and cuda" --run-level "full_model"
+          - pytest -s -v tests/e2e/online_serving/test_wan_2_1_vace_expansion.py tests/e2e/online_serving/test_ltx2_expansion.py -m "full_model and cuda" --run-level "full_model"
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Diffusion X2V · Wan2.2 Accuracy Test"
+        timeout_in_minutes: 180
+        commands:
+          - pytest -s -v tests/e2e/accuracy/wan22_i2v/test_wan22_i2v_video_similarity.py -m full_model --run-level full_model
+          - pytest -s -v tests/e2e/accuracy/test_diffusers_backend_similarity.py -k '2v_matches_diffusers' -m full_model --run-level full_model
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Diffusion X2V · Other Accuracy Test"
+        timeout_in_minutes: 180
+        commands:
+          - pytest -s -v tests/e2e/accuracy/hunyuanvideo15_t2v/test_hunyuanvideo15_t2v_video_similarity.py -m full_model --run-level full_model
+          - pytest -s -v tests/e2e/accuracy/hunyuanvideo15_i2v/test_hunyuanvideo15_i2v_video_similarity.py -m full_model --run-level full_model
+          - pytest -s -v 'tests/e2e/accuracy/ltx/test_ltx_official_similarity.py::test_ltx_one_stage_matches_official[ltx2_3_i2v]' -m full_model --run-level full_model
+          - pytest -s -v tests/e2e/accuracy/test_video_streaming_output_similarity.py -m full_model --run-level full_model
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Diffusion X2V · Perf Test"
+        key: nightly-diffusion-x2v-performance
+        timeout_in_minutes: 180
+        commands:
+          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
+          - export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py --test-config-file tests/dfx/perf/tests/test_wan22_i2v_vllm_omni.json
+            EXIT1=$$?
+            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py --test-config-file tests/dfx/perf/tests/test_cosmos3_vllm_omni.json
+            EXIT2=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
+            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
+            if [ $$EXIT1 -ne 0 ]; then exit $$EXIT1; fi
+            exit $$EXIT2
+        mirror_hardwares: h100_2
+
+  - group: ":card_index_dividers: Diffusion Test"
+    depends_on: upload-nightly-pipeline
+    if: build.env("NIGHTLY") == "1" || build.pull_request.labels includes "nightly-test" || build.pull_request.labels includes "diffusion-x2iat-test"
+    steps:
+
+      - label: ":full_moon: Diffusion · Offloader Test"
+        timeout_in_minutes: 40
+        commands:
+          - timeout 40m pytest -s -v tests/diffusion/offloader -m "full_model and cuda"
+        mirror_hardwares: l4_1
+
+      - label: ":full_moon: Diffusion · Quantization Test with H100"
+        commands:
+          - timeout 60m pytest -sv tests/diffusion/quantization -m 'full_model and cuda and H100' --run-level "full_model"
+        mirror_hardwares: h100_1
+
+      - label: ":full_moon: Diffusion · Quantization Test with L4"
+        timeout_in_minutes: 60
+        commands:
+          - pytest -sv tests/diffusion/quantization -m 'full_model and cuda and L4' --run-level "full_model"
+        mirror_hardwares: l4_1
+
+      - label: ":full_moon: Diffusion · Wan VAE Spatial Shard Correctness Test"
+        timeout_in_minutes: 60
+        commands:
+          - >-
+            pytest -sv tests/diffusion/distributed/test_wan_spatial_shard.py
+            -m "full_model and diffusion and H100 and distributed_cuda"
+            --run-level "full_model"
+        mirror_hardwares: h100_2
+
+  - label: ":bar_chart: Testcase Statistics"
+    key: nightly-testcase-statistics
+    timeout_in_minutes: 120
+    depends_on: upload-nightly-pipeline
+    if: build.env("NIGHTLY") == "1"
+    commands:
+      - python tools/nightly/buildkite_testcase_statistics.py -o tests/dfx/perf/results/buildkite_testcase_statistics.html
+      - buildkite-agent artifact upload "tests/dfx/perf/results/*.html"
+    mirror_hardwares: h100_1
+
+  # NOTE: ":email: Nightly Collection & Email" is deprecated and will be removed soon.
+  # Do not add new depends_on entries or extend this step.
+  - label: ":email: Nightly Collection & Email"
+    key: nightly-perf-distribution
+    depends_on:
+      - nightly-omni-performance-no-async-chunk
+      - nightly-omni-performance-async-chunk
+      - nightly-omni-performance-vllm-text
+      - nightly-omni-performance-multi-replicas
+      - nightly-tts-performance
+      - nightly-tts-performance-higgs-audio-v3
+      - nightly-diffusion-x2iat-performance-qwen-image
+      - nightly-diffusion-x2iat-performance-qwen-image-edit
+      - nightly-diffusion-x2iat-performance-qwen-image-edit-2511
+      - nightly-diffusion-x2iat-performance-qwen-image-layered
+      - nightly-diffusion-x2iat-performance-bagel
+      - nightly-diffusion-x2v-performance
+      - nightly-testcase-statistics
+    if: build.env("NIGHTLY") == "1" && build.env("EMAIL_DISTRIBUTION") == "1"
+    commands:
+      - pip install openpyxl
+      - export DEFAULT_INPUT_DIR=tests/dfx/perf/results
+      - export DEFAULT_OUTPUT_DIR=tests/dfx/perf/results
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-tts-performance
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-tts-performance-higgs-audio-v3
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-omni-performance-no-async-chunk
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-omni-performance-async-chunk
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-omni-performance-vllm-text
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-omni-performance-multi-replicas
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-diffusion-x2iat-performance-qwen-image
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-diffusion-x2iat-performance-qwen-image-edit
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-diffusion-x2iat-performance-qwen-image-edit-2511
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-diffusion-x2iat-performance-qwen-image-layered
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-diffusion-x2iat-performance-bagel
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-diffusion-x2v-performance
+      - buildkite-agent artifact download "tests/dfx/perf/results/*.html" . --step nightly-testcase-statistics
+      - python tools/nightly/generate_nightly_perf_excel.py
+      - python tools/nightly/generate_nightly_perf_html.py
+      - python tools/nightly/send_nightly_email.py --report-file "tests/dfx/perf/results/*.xlsx, tests/dfx/perf/results/*.html"
+      - buildkite-agent artifact upload "tests/dfx/perf/results/*.xlsx"
+      - buildkite-agent artifact upload "tests/dfx/perf/results/*.html"
+    agents:
+      queue: "cpu_queue_premerge"

.buildkite/cuda/test-ready.yml

@@ -0,0 +1,356 @@
+env:
+  VLLM_WORKER_MULTIPROC_METHOD: spawn
+  VLLM_USE_FLASHINFER_MOE_FP16: 0
+  HF_HUB_DOWNLOAD_TIMEOUT: 300
+  HF_HUB_ETAG_TIMEOUT: 60
+
+steps:
+  - group: ":card_index_dividers: Simple Test"
+    depends_on: upload-ready-pipeline
+    steps:
+      - label: "Simple · Diffusion & Model Executor Test"
+        commands:
+          - "timeout 20m pytest -sv tests/diffusion tests/model_executor -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+        mirror_hardwares: l4_1
+
+      - label: "Simple · Other Test"
+        commands:
+          - "timeout 20m pytest -sv -m 'core_model and cpu' --ignore=tests/diffusion --ignore=tests/model_executor --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+        mirror_hardwares: l4_1
+
+  - label: "Custom Pipeline Test"
+    depends_on: upload-ready-pipeline
+    commands:
+      - timeout 20m pytest -s -v tests/e2e/offline_inference/custom_pipeline/ -m "core_model"
+    mirror_hardwares: l4_1
+
+  - group: ":card_index_dividers: Diffusion Test"
+    depends_on: upload-ready-pipeline
+    steps:
+
+      - label: "Diffusion · Offloader Test"
+        commands:
+          - timeout 40m pytest -s -v tests/diffusion/offloader -m "core_model and cuda" --run-level "core_model"
+        mirror_hardwares: l4_1
+
+      - label: "Diffusion · Cache Test"
+        commands:
+          - timeout 15m pytest -sv tests/diffusion/cache -m 'core_model and cuda' --run-level "core_model"
+        mirror_hardwares: l4_1
+
+      - label: "Diffusion · Batch Test"
+        commands:
+          - timeout 15m pytest -sv tests/diffusion/batching -m 'core_model and cuda' --run-level "core_model"
+        mirror_hardwares: l4_1
+
+      - label: "Diffusion · Distributed Test"
+        source_file_dependencies:
+          - tests/diffusion/distributed/
+          - vllm_omni/diffusion/distributed/
+          - vllm_omni/diffusion/attention/parallel/
+          - vllm_omni/diffusion/data.py
+          - vllm_omni/diffusion/models/
+        commands:
+          - timeout 20m pytest -sv tests/diffusion/distributed -m 'core_model and cuda' --run-level "core_model"
+        mirror_hardwares: l4_4
+
+      - label: "Diffusion · Model Test"
+        commands:
+          - timeout 15m pytest -sv tests/diffusion/models/ -m 'core_model and cuda' --run-level "core_model"
+        mirror_hardwares: l4_1
+
+      - label: "Diffusion · Other Test"
+        commands:
+          - timeout 15m pytest -sv tests/diffusion/*.py -m 'core_model and cuda' --run-level "core_model"
+        mirror_hardwares: l4_1
+
+  - group: ":card_index_dividers: Tiny Model Tests [Single GPU]"
+    depends_on: upload-ready-pipeline
+    steps:
+
+      - label: "Tiny Model Tests · Diffusion (base)"
+        commands:
+          - timeout 15m pytest -sv tests/model_tests/diffusion -m 'core_model and cuda' --run-level "core_model"
+        mirror_hardwares: l4_1
+
+  - label: "Engine&Model Executor Test"
+    depends_on: upload-ready-pipeline
+    commands:
+      - |
+        timeout 15m bash -c "
+                  pytest -sv tests/engine/ tests/model_executor/ -m 'core_model and cuda' --run-level 'core_model'
+        "
+    mirror_hardwares: l4_1
+
+  - label: "Distributed Test"
+    depends_on: upload-ready-pipeline
+    commands:
+      - |
+        timeout 20m bash -c "
+          pytest -sv tests/distributed/ -m 'core_model and cuda and L4' --run-level 'core_model'
+        "
+    mirror_hardwares: l4_1
+
+  - label: "Entrypoints Test"
+    depends_on: upload-ready-pipeline
+    commands:
+      - |
+        timeout 30m bash -c "
+          export VLLM_WORKER_MULTIPROC_METHOD=spawn
+          pytest -sv tests/entrypoints/ -m 'core_model and cuda' --run-level 'core_model'
+        "
+    mirror_hardwares: l4_1
+
+  - group: ":card_index_dividers: E2E Test"
+    depends_on: upload-ready-pipeline
+    steps:
+
+      - label: "RLHF · VeRL-Omni E2E Test"
+        source_file_dependencies:
+          - tests/e2e/offline_inference/rlhf_test/test_verl_omni_e2e.py
+          - tests/e2e/offline_inference/custom_pipeline/
+          - vllm_omni/diffusion/
+          - vllm_omni/engine/
+          - vllm_omni/entrypoints/
+          - vllm_omni/inputs/
+          - vllm_omni/outputs/
+        commands:
+          - |
+            timeout 20m bash -c "
+              pytest -s -v tests/e2e/offline_inference/rlhf_test/test_verl_omni_e2e.py
+            "
+        mirror_hardwares: l4_1
+
+      - label: "Omni · Qwen3-Omni Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen3_omni.py
+          - vllm_omni/model_executor/models/qwen3_omni/
+          - vllm_omni/model_executor/stage_input_processors/qwen3_omni.py
+          - vllm_omni/deploy/qwen3_omni_moe.yaml
+        commands:
+          - |
+            timeout 20m bash -c "
+              pytest -s -v tests/e2e/online_serving/test_qwen3_omni.py -m 'core_model' --run-level 'core_model'
+            "
+        mirror_hardwares: h100_2
+
+      - label: "Omni · MiniCPM-o 4.5 Offline Test"
+        source_file_dependencies:
+          - tests/e2e/offline_inference/test_minicpmo_4_5.py
+          - vllm_omni/deploy/minicpmo_4_5.yaml
+        commands:
+          - |
+            timeout 20m bash -c "
+              pytest -s -v tests/e2e/offline_inference/test_minicpmo_4_5.py -m 'core_model' --run-level 'core_model'
+            "
+        agents:
+          queue: "mithril-h100-pool"
+        plugins:
+          - kubernetes:
+              podSpec:
+                containers:
+                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
+                    resources:
+                      limits:
+                        nvidia.com/gpu: 2
+                    volumeMounts:
+                      - name: devshm
+                        mountPath: /dev/shm
+                      - name: hf-cache
+                        mountPath: /root/.cache/huggingface
+                    env:
+                      - name: HF_HOME
+                        value: /root/.cache/huggingface
+                nodeSelector:
+                  node.kubernetes.io/instance-type: gpu-h100-sxm
+                volumes:
+                  - name: devshm
+                    emptyDir:
+                      medium: Memory
+                  - name: hf-cache
+                    hostPath:
+                      path: /mnt/hf-cache
+                      type: DirectoryOrCreate
+
+      - label: "Omni · MiniCPM-o 4.5 Online Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_minicpmo_4_5.py
+          - vllm_omni/deploy/minicpmo_4_5.yaml
+        commands:
+          - |
+            timeout 20m bash -c "
+              pytest -s -v tests/e2e/online_serving/test_minicpmo_4_5.py -m 'core_model' --run-level 'core_model'
+            "
+        agents:
+          queue: "mithril-h100-pool"
+        plugins:
+          - kubernetes:
+              podSpec:
+                containers:
+                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
+                    resources:
+                      limits:
+                        nvidia.com/gpu: 2
+                    volumeMounts:
+                      - name: devshm
+                        mountPath: /dev/shm
+                      - name: hf-cache
+                        mountPath: /root/.cache/huggingface
+                    env:
+                      - name: HF_HOME
+                        value: /root/.cache/huggingface
+                nodeSelector:
+                  node.kubernetes.io/instance-type: gpu-h100-sxm
+                volumes:
+                  - name: devshm
+                    emptyDir:
+                      medium: Memory
+                  - name: hf-cache
+                    hostPath:
+                      path: /mnt/hf-cache
+                      type: DirectoryOrCreate
+
+      - label: "TTS · VoxCPM2 Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_voxcpm2_tts.py
+          - vllm_omni/model_executor/models/voxcpm2/
+          - vllm_omni/deploy/voxcpm2.yaml
+        timeout_in_minutes: 30
+        commands:
+          - |
+            timeout 30m bash -c "
+              export VLLM_LOGGING_LEVEL=DEBUG
+              export VLLM_WORKER_MULTIPROC_METHOD=spawn
+              pytest -s -v tests/e2e/online_serving/test_voxcpm2_tts.py -m 'core_model' --run-level 'core_model'
+            "
+        mirror_hardwares: l4_1
+
+      - label: "TTS · Qwen3-TTS CustomVoice Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen3_tts_customvoice.py
+          - vllm_omni/model_executor/models/qwen3_tts/
+          - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
+          - vllm_omni/deploy/qwen3_tts.yaml
+        commands:
+          - |
+            timeout 20m bash -c "
+              export VLLM_ALLOW_LONG_MAX_MODEL_LEN=1
+              pytest -s -v tests/e2e/online_serving/test_qwen3_tts_customvoice.py -m 'core_model' --run-level 'core_model'
+            "
+        mirror_hardwares: l4_1
+
+      - label: "TTS · Qwen3-TTS Base Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen3_tts_base.py
+          - vllm_omni/model_executor/models/qwen3_tts/
+          - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
+          - vllm_omni/deploy/qwen3_tts.yaml
+        commands:
+          - |
+            timeout 20m bash -c "
+              export VLLM_WORKER_MULTIPROC_METHOD=spawn
+              export VLLM_ALLOW_LONG_MAX_MODEL_LEN=1
+              export VLLM_OMNI_USE_V2_RUNNER=1
+              pytest -s -v tests/e2e/online_serving/test_qwen3_tts_base.py -m 'core_model' --run-level 'core_model'
+            "
+        mirror_hardwares: l4_1
+
+      - label: "TTS · Higgs-Audio-V3 Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_higgs_audio_v3.py
+          - vllm_omni/model_executor/models/higgs_audio_v3/
+          - vllm_omni/model_executor/stage_input_processors/higgs_audio_v3.py
+          - vllm_omni/deploy/higgs_multimodal_qwen3.yaml
+        commands:
+          - |
+            timeout 20m bash -c "
+              export VLLM_WORKER_MULTIPROC_METHOD=spawn
+              export VLLM_USE_DEEP_GEMM=0
+              export VLLM_MOE_USE_DEEP_GEMM=0
+              pytest -s -v tests/e2e/online_serving/test_higgs_audio_v3.py -m 'core_model' --run-level 'core_model'
+            "
+        mirror_hardwares: l4_1
+
+      - label: "Diffusion · Qwen Image Edit Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen_image_edit.py
+          - vllm_omni/diffusion/models/qwen_image/
+        commands:
+          - |
+            timeout 20m bash -c "
+              pytest -s -v tests/e2e/online_serving/test_qwen_image_edit.py -m 'core_model' --run-level 'core_model'
+            "
+        mirror_hardwares: h100_1
+
+      - label: "Diffusion · Bagel Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_bagel.py
+          - vllm_omni/model_executor/models/bagel/
+          - vllm_omni/diffusion/models/bagel/
+          - vllm_omni/model_executor/stage_input_processors/bagel.py
+          - vllm_omni/deploy/bagel.yaml
+        commands:
+          - |
+            timeout 40m bash -c "
+              export VLLM_IMAGE_FETCH_TIMEOUT=60
+              pytest -s -v tests/e2e/online_serving/test_bagel.py -m 'core_model' --run-level 'core_model'
+            "
+        mirror_hardwares: h100_1
+
+      - label: "Diffusion · Wan22 Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_wan22_t2v.py
+          - vllm_omni/diffusion/models/wan2_2/
+        commands:
+          - |
+            timeout 40m bash -c "
+              export VLLM_IMAGE_FETCH_TIMEOUT=60
+              pytest -s -v tests/e2e/online_serving/test_wan22_t2v.py -m 'core_model' --run-level 'core_model'
+            "
+        mirror_hardwares: h100_1
+
+      - label: "Diffusion · Cosmos3 Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_cosmos3.py
+          - vllm_omni/diffusion/models/cosmos3/
+        commands:
+          - |
+            timeout 40m bash -c "
+              export VLLM_IMAGE_FETCH_TIMEOUT=60
+              pytest -s -v tests/e2e/online_serving/test_cosmos3.py -m 'core_model' --run-level 'core_model'
+            "
+        mirror_hardwares: h100_1
+
+      - label: "Diffusion · Qwen Image Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen_image.py
+          - vllm_omni/diffusion/models/qwen_image/
+        commands:
+          - |
+            timeout 40m bash -c "
+              export VLLM_IMAGE_FETCH_TIMEOUT=60
+              pytest -s -v tests/e2e/online_serving/test_qwen_image.py -m 'core_model' --run-level 'core_model'
+            "
+        mirror_hardwares: h100_1
+
+      - label: "Diffusion · Qwen Image Layered Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen_image_layered.py
+          - vllm_omni/diffusion/models/qwen_image/
+        commands:
+          - |
+            timeout 40m bash -c "
+              export VLLM_IMAGE_FETCH_TIMEOUT=60
+              pytest -s -v tests/e2e/online_serving/test_qwen_image_layered.py -m 'core_model' --run-level 'core_model'
+            "
+        mirror_hardwares: h100_1
+
+      - label: "Diffusion · GGUF Plugin Test"
+        soft_fail: true
+        source_file_dependencies:
+          - tests/plugin/test_gguf_diffusion.py
+          - vllm_omni/diffusion/model_loader
+        commands:
+          - pip install "vllm-gguf-plugin>=0.0.3"
+          - timeout 30m pytest -sv tests/plugin/test_gguf_diffusion.py
+        mirror_hardwares: l4_1

.buildkite/cuda/test-weekly.yml

@@ -0,0 +1,136 @@
+env:
+  VLLM_WORKER_MULTIPROC_METHOD: spawn
+  VLLM_USE_FLASHINFER_MOE_FP16: 0
+  HF_HUB_DOWNLOAD_TIMEOUT: 300
+  HF_HUB_ETAG_TIMEOUT: 60
+
+steps:
+
+  - label: "Reliability Test - qwen3-omni"
+    timeout_in_minutes: 180
+    depends_on: upload-weekly-pipeline
+    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
+    commands:
+      - pytest -s -v tests/dfx/reliability/test_reliability_qwen3_omni.py -m "slow"
+    mirror_hardwares: h100_2
+
+  - label: "Reliability Test - wan22"
+    timeout_in_minutes: 180
+    depends_on: upload-weekly-pipeline
+    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
+    commands:
+      - pytest -s -v tests/dfx/reliability/test_reliability_wan22.py -m "slow"
+    mirror_hardwares: h100_2
+
+  - label: "Reliability Test - hunyuan-image"
+    timeout_in_minutes: 180
+    depends_on: upload-weekly-pipeline
+    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
+    commands:
+      - pytest -s -v tests/dfx/reliability/test_reliability_hunyuan_image.py -m "slow"
+    mirror_hardwares: h100_4
+
+  - label: "Reliability Test - voxcpm2"
+    timeout_in_minutes: 180
+    depends_on: upload-weekly-pipeline
+    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
+    commands:
+      - pytest -s -v tests/dfx/reliability/test_reliability_voxcpm2.py -m "slow"
+    mirror_hardwares: l4_1
+
+  - group: "Reliability Test - Invalid parameters Test"
+    depends_on: upload-weekly-pipeline
+    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
+    steps:
+
+      - label: "Invalid parameters Test · H100"
+        timeout_in_minutes: 120
+        commands:
+          - pytest -s -v tests/dfx/reliability/invalid_param_test/ -m "slow and H100"
+        mirror_hardwares: h100_2
+
+      - label: "Invalid parameters Test · L4"
+        timeout_in_minutes: 120
+        commands:
+          - pytest -s -v tests/dfx/reliability/invalid_param_test/ -m "slow and L4"
+        mirror_hardwares: l4_1
+
+  - group: ":card_index_dividers: E2E Tests"
+    depends_on: upload-weekly-pipeline
+    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
+    steps:
+
+      - label: "Omni · L4"
+        timeout_in_minutes: 90
+        commands:
+          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
+          - pytest -s -v tests/e2e/*/test_qwen2_5_omni_expansion.py --run-level "core_model"
+          - >-
+            pytest -s -v tests/e2e/
+            --ignore=tests/e2e/online_serving/test_qwen2_5_omni_expansion.py
+            --ignore=tests/e2e/offline_inference/test_qwen2_5_omni_expansion.py
+            -m "slow and L4 and omni"
+            --run-level "full_model"
+        mirror_hardwares: l4_4
+
+      - label: "Omni · H100"
+        timeout_in_minutes: 90
+        commands:
+          - >-
+            pytest -sv
+            tests/e2e/ -m "slow and H100 and omni"
+            --run-level "full_model"
+        mirror_hardwares: h100_2
+
+      - label: "TTS · L4"
+        timeout_in_minutes: 90
+        commands:
+          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
+          - export VLLM_USE_DEEP_GEMM="0"
+          - export VLLM_MOE_USE_DEEP_GEMM="0"
+          - |
+            failed=0
+            for f in $(pytest --collect-only -q tests/e2e/ -m "slow and L4 and tts" --run-level "full_model" | grep '\.py::' | sed -E 's|::.*||' | sort -u); do
+              pytest -s -v "$$f" -m "slow and L4 and tts" --run-level "full_model" || failed=1
+            done
+            exit $$failed
+        mirror_hardwares: l4_1
+
+      - label: "TTS · H100"
+        timeout_in_minutes: 90
+        commands:
+          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
+          - export VLLM_USE_DEEP_GEMM="0"
+          - export VLLM_MOE_USE_DEEP_GEMM="0"
+          - >-
+            pytest -sv
+            tests/e2e/ -m "slow and H100 and tts"
+            --run-level "full_model"
+        mirror_hardwares: h100_1
+
+      - label: "Diffusion · L4"
+        timeout_in_minutes: 60
+        commands:
+          - >-
+            pytest -sv
+            tests/e2e/ -m "slow and diffusion and L4"
+            --run-level "full_model"
+        mirror_hardwares: l4_4
+
+      - label: "Diffusion · H100 · Single-GPU"
+        timeout_in_minutes: 120
+        commands:
+          - >-
+            pytest -sv
+            tests/e2e/ -m "slow and diffusion and H100 and not distributed_cuda"
+            --run-level "full_model"
+        mirror_hardwares: h100_1
+
+      - label: "Diffusion · H100 · Multi-GPU"
+        timeout_in_minutes: 90
+        commands:
+          - >-
+            pytest -sv
+            tests/e2e/ -m "slow and diffusion and H100 and distributed_cuda"
+            --run-level "full_model"
+        mirror_hardwares: h100_2

.buildkite/intel/pipeline-intel.yml

@@ -1,9 +1,9 @@
-# .buildkite/pipeline-intel.yaml
+# .buildkite/intel/pipeline-intel.yml
 steps:
   - group: "INTEL Tests"
     steps:
       - label: "Intel: XPU Test"
-        command: "bash .buildkite/scripts/hardware_ci/run-xpu-test.sh"
+        command: "bash .buildkite/intel/scripts/run-xpu-test.sh"
         agents:
           queue: "intel-gpu-omni"
         env:

.buildkite/intel/scripts/bootstrap-intel-omni.sh

@@ -0,0 +1,16 @@
+#!/bin/bash
+# vllm-omni Intel bootstrap
+# Uses static pipeline-intel.yml for Intel XPU tests
+
+set -euo pipefail
+
+source .buildkite/common/scripts/resolve_skip_ci.sh
+
+upload_pipeline() {
+    echo "--- 🛠 Preparing Intel pipeline"
+    buildkite-agent pipeline upload .buildkite/intel/pipeline-intel.yml
+}
+
+gate_bootstrap_ci intel l2
+
+upload_pipeline

.buildkite/intel/scripts/run-xpu-test.sh

@@ -74,6 +74,6 @@ time timeout -k 30 30m docker run \
     cd /workspace/vllm-omni
     pytest -v -s -m "core_model and xpu and B60"
     pytest -v -s tests/diffusion/quantization/test_mxfp8_config.py
-    export VLLM_XPU_USE_SAMPLER_KERNEL=0    # NOTE: Remove this after vLLM v0.21.1 is merged. Fixes Qwen2-5 omni-expansion tests.
     pytest -v -s -m "advanced_model and xpu and B60"
+    pytest -v -s -m "omni and xpu and B60"
 '

.buildkite/npu/bootstrap-upload-steps.yml

@@ -1,6 +1,8 @@
+# Uploaded by upload_pipeline.py --upload (not the Buildkite hook).
+# ``if`` is injected by step ``key`` from skip-ci at upload time.
 steps:
   - label: ":buildkit: Build and Push NPU Test Image - A2"
-    key: image-build
+    key: image-build-a2
     agents:
       queue: "ascend-a2b3"
       resource_class: "npu-1"
@@ -159,37 +161,21 @@ steps:
 
   - label: "Upload Ready Pipeline"
     depends_on:
-      - image-build
+      - image-build-a2
       - image-build-a3
     key: upload-ready-pipeline
-    if: >-
-      build.branch != "main" && (
-        build.pull_request.labels includes "npu-test" ||
-        build.pull_request.labels includes "ready"
-      )
     agents:
-      queue: "ascend-a2b3"
-      resource_class: "npu-1"
+      queue: "cpu_queue_premerge"
     command: |
-      buildkite-agent pipeline upload .buildkite/test-npu-ready.yaml
+      python3 .buildkite/common/scripts/upload_pipeline.py --upload .buildkite/npu/test-npu-ready.yml
 
   # L4 Test — main+NIGHTLY=1 (scheduled), or PR with specific label (e.g. add label then Rebuild)
   - label: "Upload Nightly Pipeline"
     depends_on:
-      - image-build
+      - image-build-a2
       - image-build-a3
     key: upload-nightly-pipeline
-    if: >-
-      (build.branch == "main" && build.env("NIGHTLY") == "1") ||
-      (build.branch != "main" && (
-        build.pull_request.labels includes "nightly-test" ||
-        build.pull_request.labels includes "omni-test" ||
-        build.pull_request.labels includes "tts-test" ||
-        build.pull_request.labels includes "diffusion-x2iat-test" ||
-        build.pull_request.labels includes "diffusion-x2v-test"
-      ))
     commands:
-      - buildkite-agent pipeline upload .buildkite/test-npu-nightly.yaml
+      - python3 .buildkite/common/scripts/upload_pipeline.py --upload .buildkite/npu/test-npu-nightly.yml
     agents:
-      queue: "ascend-a2b3"
-      resource_class: "npu-1"
+      queue: "cpu_queue_premerge"

.buildkite/npu/pipeline-npu-a3.yml

@@ -88,7 +88,7 @@
 #   #     queue: "ascend-a3"
 #   #     resource_class: "npu-2"
 #   #   command: |
-#   #     buildkite-agent pipeline upload .buildkite/test-npu-ready.yaml
+#   #     buildkite-agent pipeline upload .buildkite/npu/test-npu-ready.yml
 
   # # L4 Test — main+NIGHTLY=1 (scheduled), or PR with specific label (e.g. add label then Rebuild)
   # - label: "Upload Nightly Pipeline"
@@ -104,7 +104,7 @@
   #       build.pull_request.labels includes "diffusion-x2v-test"
   #     ))
   #   commands:
-  #     - buildkite-agent pipeline upload .buildkite/test-npu-nightly.yaml
+  #     - buildkite-agent pipeline upload .buildkite/npu/test-npu-nightly.yml
   #   agents:
   #     queue: "ascend-a3"
   #     resource_class: "npu-2"

.buildkite/npu/pipeline-npu.yml

@@ -0,0 +1,9 @@
+# Hook uploads this entry YAML; upload_pipeline.py --upload reads bootstrap-upload-steps.yml
+# and injects ``if`` by step ``key`` from skip-ci.
+steps:
+  - label: ":github: Resolve skip-ci & upload pipeline"
+    key: upload-ci-pipeline
+    commands:
+      - "python3 .buildkite/common/scripts/upload_pipeline.py --upload .buildkite/npu/bootstrap-upload-steps.yml"
+    agents:
+      queue: "cpu_queue_premerge"

.buildkite/npu/test-npu-nightly.yml

@@ -0,0 +1,123 @@
+steps:
+  - group: ":card_index_dividers: Omni Model Test"
+    key: nightly-omni-test-group
+    depends_on: upload-nightly-pipeline
+    if: >-
+      build.env("NIGHTLY") == "1" ||
+      build.pull_request.labels includes "nightly-test" ||
+      build.pull_request.labels includes "omni-test"
+    steps:
+      - label: ":full_moon: Omni · Perf Test · No Async Chunk"
+        key: nightly-omni-performance-no-async-chunk
+        timeout_in_minutes: 180
+        mirror_hardwares: a3_npu_4
+        env:
+          VLLM_WORKER_MULTIPROC_METHOD: spawn
+          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
+          HF_TOKEN: "${HF_TOKEN}"
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py -m npu --test-config-file tests/dfx/perf/tests/test_qwen3_omni_no_async_chunk.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+      - label: ":full_moon: Omni · Perf Test · Async Chunk"
+        key: nightly-omni-performance-async-chunk
+        timeout_in_minutes: 180
+        mirror_hardwares: a3_npu_4
+        env:
+          VLLM_WORKER_MULTIPROC_METHOD: spawn
+          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
+          HF_TOKEN: "${HF_TOKEN}"
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py -m npu --test-config-file tests/dfx/perf/tests/test_qwen3_omni_async_chunk.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+      - label: ":robot: Omni · MiniCPM-o 4.5 Test"
+        mirror_hardwares: a3_npu_2
+        env:
+          VLLM_WORKER_MULTIPROC_METHOD: spawn
+          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
+          HF_TOKEN: "${HF_TOKEN}"
+        commands:
+          - pytest -s -v tests/e2e/offline_inference/test_minicpmo_4_5.py tests/e2e/online_serving/test_minicpmo_4_5.py tests/e2e/online_serving/test_minicpmo_4_5_expansion.py -m 'full_model' --run-level 'full_model'
+  - group: ":card_index_dividers: TTS Model Test"
+    key: nightly-tts-test-group
+    depends_on: upload-nightly-pipeline
+    if: build.env("NIGHTLY") == "1" || build.pull_request.labels includes "nightly-test" || build.pull_request.labels includes "tts-test"
+    steps:
+      - label: ":full_moon: TTS · Perf Test"
+        timeout_in_minutes: 180
+        mirror_hardwares: a3_npu_2
+        env:
+          VLLM_WORKER_MULTIPROC_METHOD: spawn
+          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
+          HF_TOKEN: "${HF_TOKEN}"
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py -m npu --test-config-file tests/dfx/perf/tests/test_tts.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+  - group: ":card_index_dividers: Diffusion X2I(&A&T) Model Test"
+    key: nightly-diffusion-x2iat-group
+    depends_on: upload-nightly-pipeline
+    if: >-
+      build.env("NIGHTLY") == "1" ||
+      build.pull_request.labels includes "nightly-test" ||
+      build.pull_request.labels includes "diffusion-x2iat-test"
+    steps:
+      - label: ":full_moon: Diffusion X2I(&A&T) · Perf Test · Qwen-Image-Edit-2511"
+        key: nightly-diffusion-x2iat-performance-qwen-image-edit-2511
+        timeout_in_minutes: 180
+        mirror_hardwares: a3_npu_8
+        env:
+          VLLM_WORKER_MULTIPROC_METHOD: spawn
+          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
+          HF_TOKEN: "${HF_TOKEN}"
+          DIFFUSION_ATTENTION_BACKEND: TORCH_SDPA
+        commands:
+          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
+          - export CACHE_DIT_VERSION=1.3.0
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py -m npu --test-config-file tests/dfx/perf/tests/test_qwen_image_edit_2511_vllm_omni.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
+            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
+            exit $$EXIT
+  - group: ":card_index_dividers: Diffusion X2V Model Test"
+    key: nightly-diffusion-x2v-group
+    depends_on: upload-nightly-pipeline
+    if: >-
+      build.env("NIGHTLY") == "1" ||
+      build.pull_request.labels includes "nightly-test" ||
+      build.pull_request.labels includes "diffusion-x2v-test"
+    steps:
+      - label: ":full_moon: Diffusion X2V · Perf Test"
+        key: nightly-diffusion-x2v-performance
+        timeout_in_minutes: 180
+        mirror_hardwares: a3_npu_4
+        env:
+          VLLM_WORKER_MULTIPROC_METHOD: spawn
+          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
+          HF_TOKEN: "${HF_TOKEN}"
+          DIFFUSION_ATTENTION_BACKEND: TORCH_SDPA
+        commands:
+          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py -m npu --test-config-file tests/dfx/perf/tests/test_wan22_i2v_vllm_omni.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
+            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
+            exit $$EXIT

.buildkite/npu/test-npu-ready.yml

@@ -0,0 +1,34 @@
+steps:
+  - group: ":card_index_dividers: Omni Model Test"
+    key: ready-omni-group
+    depends_on: upload-ready-pipeline
+    steps:
+      - label: ":robot: Omni · MiniCPM-o 4.5 Test"
+        agents:
+          queue: "ascend-a3"
+          resource_class: "npu-2"
+        image: "swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/vllm-omni-ci-npu:${BUILDKITE_COMMIT}-a3"
+        plugins:
+          - kubernetes:
+              podSpecPatch:
+                imagePullSecrets:
+                  - name: swr-secret
+        env:
+          VLLM_WORKER_MULTIPROC_METHOD: spawn
+          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
+          HF_TOKEN: "${HF_TOKEN}"
+        commands:
+          - pytest -s -v tests/e2e/offline_inference/test_minicpmo_4_5.py tests/e2e/online_serving/test_minicpmo_4_5.py -m 'core_model' --run-level 'core_model'
+
+  - group: ":card_index_dividers: TTS Model Test"
+    key: ready-tts-group
+    depends_on: upload-ready-pipeline
+    steps:
+      - label: ":sound: TTS · Qwen3-TTS CustomVoice Test"
+        mirror_hardwares: a2b3_npu_1
+        env:
+          VLLM_WORKER_MULTIPROC_METHOD: spawn
+          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
+          HF_TOKEN: "${HF_TOKEN}"
+        commands:
+          - pytest -s -v tests/e2e/online_serving/test_qwen3_tts_customvoice.py -m 'core_model and npu' --run-level 'core_model'

.buildkite/pipeline.yml

@@ -1,87 +0,0 @@
-# Document 1: Buildkite loads only this block on first parse. The next step resolves docs-only,
-# pytest skip-mark-only, or combined skip-ci from git diff, then uploads document 2. When skip-ci
-# applies, image-build still runs for the nightly exception in upload_pipeline.py (nightly-test
-# label or main NIGHTLY=1); otherwise upload-nightly can be skipped together with test-ready/test-merge.
-# Weekly uploads also depend on image-build.
-
-# Document 2: appended after `---`; same file, read by upload_pipeline.py (not evaluated as a second pipeline by Buildkite). Child uploads include test-ready.yml, test-merge.yml, test-nightly.yml, and test-weekly.yml.
-steps:
-  - label: ":github: Resolve skip-ci (docs / skip marks) & upload pipeline"
-    key: upload-ci-pipeline
-    commands:
-      - "python3 .buildkite/scripts/upload_pipeline.py --upload .buildkite/pipeline.yml"
-    agents:
-      queue: "cpu_queue_premerge"
-
----
-steps:
-  - label: ":docker: Build image"
-    key: image-build
-    if: __IMAGE_BUILD_IF__
-    commands:
-      - "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7"
-      - "docker build --progress=plain --file docker/Dockerfile.ci -t vllm-omni-ci ."
-      - "docker tag vllm-omni-ci public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT"
-      - "docker push public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT"
-    agents:
-      queue: "cpu_queue_premerge"
-
-  # L2 Test — PR ready label (diff-aware) or main+NIGHTLY=1 (--e2e, full E2E group)
-  - label: "Upload Ready Pipeline"
-    depends_on: image-build
-    key: upload-ready-pipeline
-    if: __UPLOAD_READY_IF__
-    commands:
-      - |
-        EXTRA=""
-        if [ "${NIGHTLY:-}" = "1" ] && [ "${BUILDKITE_BRANCH}" = "main" ]; then
-          EXTRA="--e2e"
-        fi
-        python3 .buildkite/scripts/upload_pipeline.py --upload $$EXTRA .buildkite/test-ready.yml
-    agents:
-      queue: "cpu_queue_premerge"
-
-  # L3 Test — PR merge-test label (diff-aware) or main+NIGHTLY=1 (--e2e, full E2E group)
-  - label: "Upload Merge Pipeline"
-    depends_on: image-build
-    key: upload-merge-pipeline
-    if: __UPLOAD_MERGE_IF__
-    commands:
-      - |
-        EXTRA=""
-        if [ "${NIGHTLY:-}" = "1" ] && [ "${BUILDKITE_BRANCH}" = "main" ]; then
-          EXTRA="--e2e"
-        fi
-        python3 .buildkite/scripts/upload_pipeline.py --upload $$EXTRA .buildkite/test-merge.yml
-    agents:
-      queue: "cpu_queue_premerge"
-
-  # L4 Test — main+NIGHTLY=1 (scheduled), or PR with specific label (e.g. add label then Rebuild)
-  - label: "Upload Nightly Pipeline"
-    depends_on: image-build
-    key: upload-nightly-pipeline
-    if: >-
-      (build.branch == "main" && build.env("NIGHTLY") == "1") ||
-      (build.branch != "main" && (
-        build.pull_request.labels includes "nightly-test" ||
-        build.pull_request.labels includes "omni-test" ||
-        build.pull_request.labels includes "tts-test" ||
-        build.pull_request.labels includes "diffusion-x2iat-test" ||
-        build.pull_request.labels includes "diffusion-x2v-test"
-      ))
-    commands:
-      - buildkite-agent pipeline upload .buildkite/test-nightly.yml
-    agents:
-      queue: "cpu_queue_premerge"
-
-  # L5 Test — main+WEEKLY=1 (scheduled), or PR with weekly-test label
-  - label: "Upload Weekly Pipeline"
-    depends_on: image-build
-    key: upload-weekly-pipeline
-    if: >-
-      (build.branch == "main" && build.env("WEEKLY") == "1") ||
-      (build.branch != "main" && build.pull_request.labels includes "weekly-test")
-    commands:
-      - buildkite-agent pipeline upload .buildkite/test-weekly.yml
-    agents:
-      queue: "cpu_queue_premerge"

.buildkite/rebase-pipeline.yaml

@@ -1,38 +0,0 @@
-steps:
-  - label: ":docker: Build image"
-    key: image-build
-    commands:
-      - "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7"
-      - "docker build --progress=plain --file docker/Dockerfile.ci -t vllm-omni-ci ."
-      - "docker tag vllm-omni-ci public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT"
-      - "docker push public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT"
-    agents:
-      queue: "cpu_queue_premerge"
-
-  # L2 Test
-  - label: "Upload Ready Pipeline"
-    depends_on: image-build
-    key: upload-ready-pipeline
-    commands:
-      - python3 .buildkite/scripts/upload_pipeline.py --upload --all .buildkite/test-ready.yml
-    agents:
-      queue: "cpu_queue_premerge"
-
-  # L3 Test
-  - label: "Upload Merge Pipeline"
-    depends_on: image-build
-    key: upload-merge-pipeline
-    commands:
-      - python3 .buildkite/scripts/upload_pipeline.py --upload --all .buildkite/test-merge.yml
-    agents:
-      queue: "cpu_queue_premerge"
-
-  # L4 Test
-  - label: "Upload Nightly Pipeline"
-    depends_on: image-build
-    key: upload-nightly-pipeline
-    if: build.env("NIGHTLY") == "1"
-    commands:
-      - python3 .buildkite/scripts/upload_pipeline.py --upload --all .buildkite/test-nightly.yml
-    agents:
-      queue: "cpu_queue_premerge"

.buildkite/release/release-pipeline.yml

@@ -16,7 +16,7 @@ steps:
       - "uv venv --python=3.12 && source .venv/bin/activate"
       - "uv pip install --upgrade build"
       - "python3 -m build"
-      - "bash .buildkite/scripts/upload-nightly-wheels.sh"
+      - "bash .buildkite/release/scripts/upload-nightly-wheels.sh"
 
   - label: "Generate and upload wheel indices"
     depends_on: "build-wheel"
@@ -24,7 +24,7 @@ steps:
     agents:
       queue: small_cpu_queue_release
     commands:
-      - "bash .buildkite/scripts/generate-and-upload-nightly-index.sh"
+      - "bash .buildkite/release/scripts/generate-and-upload-nightly-index.sh"
 
   - block: "Upload release wheels to PyPI"
     depends_on: "build-wheel"
@@ -38,7 +38,7 @@ steps:
     agents:
       queue: small_cpu_queue_release
     commands:
-      - "bash .buildkite/scripts/publish-release-wheels-pypi.sh"
+      - "bash .buildkite/release/scripts/publish-release-wheels-pypi.sh"
 
   - block: "Unblock to build release Docker images"
     depends_on: ~
@@ -100,7 +100,7 @@ steps:
     agents:
       queue: small_cpu_queue_release
     commands:
-      - "bash .buildkite/scripts/publish-release-images.sh"
+      - "bash .buildkite/release/scripts/publish-release-images.sh"
     plugins:
       - docker-login#v3.0.0:
           username: vllmbot

.buildkite/release/scripts/generate-and-upload-nightly-index.sh

@@ -36,12 +36,12 @@ mkdir -p "$INDICES_OUTPUT_DIR"
 
 # HACK: we do not need regex module here, but it is required by pre-commit hook
 # To avoid any external dependency, we simply replace it back to the stdlib re module
-sed -i.bak 's/import regex as re/import re/g' .buildkite/scripts/generate-nightly-index.py && rm -f .buildkite/scripts/generate-nightly-index.py.bak
+sed -i.bak 's/import regex as re/import re/g' .buildkite/release/scripts/generate-nightly-index.py && rm -f .buildkite/release/scripts/generate-nightly-index.py.bak
 
 # Generate indices -- the version is just the commit hash (not omni/{commit})
 # because relative paths are computed between the index and wheel directories,
 # both of which live under the omni/ prefix in S3.
-$PYTHON .buildkite/scripts/generate-nightly-index.py \
+$PYTHON .buildkite/release/scripts/generate-nightly-index.py \
     --version "$BUILDKITE_COMMIT" \
     --current-objects "$obj_json" \
     --output-dir "$INDICES_OUTPUT_DIR" \
@@ -78,7 +78,7 @@ if [[ "$version" != *"dev"* ]]; then
     rm -rf "${INDICES_OUTPUT_DIR:?}"
     mkdir -p "$INDICES_OUTPUT_DIR"
     # wheel-dir is overridden to be the commit directory, so that the indices point to the correct wheel path
-    $PYTHON .buildkite/scripts/generate-nightly-index.py \
+    $PYTHON .buildkite/release/scripts/generate-nightly-index.py \
         --version "$s3_version" \
         --wheel-dir "$BUILDKITE_COMMIT" \
         --current-objects "$obj_json" \

.buildkite/scripts/docker_login_ecr_public.sh

@@ -1,73 +0,0 @@
-#!/bin/bash
-# Helper function to safely login to ECR Public with per-job config isolation
-# Uses DOCKER_CONFIG environment variable to prevent race conditions
-#
-# This script prevents the "device or resource busy" error by giving each
-# Buildkite job its own isolated Docker config directory.
-#
-# Usage:
-#   source docker_login_ecr_public.sh && safe_docker_login_ecr_public
-
-set -euo pipefail
-
-# Configuration
-ECR_REGISTRY="public.ecr.aws"
-
-setup_isolated_docker_config() {
-    # Use BUILDKITE_JOB_ID for job-specific isolation
-    # Fallback to PID if running outside Buildkite
-    local job_id="${BUILDKITE_JOB_ID:-$$}"
-
-    # Set Docker config to job-specific directory
-    export DOCKER_CONFIG="/tmp/docker-config-${job_id}"
-
-    # Create directory if it doesn't exist
-    mkdir -p "$DOCKER_CONFIG"
-
-    echo "[docker-config] Using isolated Docker config: $DOCKER_CONFIG"
-}
-
-check_docker_auth() {
-    # Check if already authenticated to the given registry
-    # Returns 0 if authenticated, 1 if not
-    local registry="$1"
-
-    # Check if credentials exist in the isolated config
-    if [[ -f "$DOCKER_CONFIG/config.json" ]]; then
-        # Check if registry is present in config
-        if grep -q "$registry" "$DOCKER_CONFIG/config.json" 2>/dev/null; then
-            return 0
-        fi
-    fi
-
-    return 1
-}
-
-safe_docker_login_ecr_public() {
-    # Setup isolated config first
-    setup_isolated_docker_config
-
-    local registry="$ECR_REGISTRY"
-
-    # Check if already authenticated (within this job)
-    if check_docker_auth "$registry"; then
-        echo "[docker-login] Already authenticated to $registry in this job"
-        return 0
-    fi
-
-    # Perform login to isolated config directory
-    echo "[docker-login] Logging in to $ECR_REGISTRY (isolated config)..."
-    if aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin "$ECR_REGISTRY"; then
-        echo "[docker-login] Login successful (config: $DOCKER_CONFIG)"
-        return 0
-    else
-        local exit_code=$?
-        echo "[docker-login] ERROR: Login failed with exit code $exit_code" >&2
-        return $exit_code
-    fi
-}
-
-# Execute if run as script (not sourced)
-if [[ "${BASH_SOURCE[0]}" == "${0}" ]]; then
-    safe_docker_login_ecr_public
-fi

.buildkite/scripts/hardware_ci/run_npu_test.sh

@@ -1,145 +0,0 @@
-#!/bin/bash
-
-# This script build the Ascend NPU docker image and run the offline inference inside the container.
-# It serves a sanity check for compilation and basic model usage.
-set -ex
-
-# Base ubuntu image with basic ascend development libraries and python installed
-VLLM_OMNI_REPO="https://github.com/vllm-project/vllm-omni.git"
-BASE_IMAGE_NAME="quay.nju.edu.cn/ascend/vllm-ascend:v0.11.0rc2"
-image_name="npu/vllm-omni-ci:${BUILDKITE_COMMIT}_${EPOCHSECONDS}"
-# image_name="npu/vllm-ci:${BUILDKITE_COMMIT}_${EPOCHSECONDS}"
-container_name="npu_${BUILDKITE_COMMIT}_$(tr -dc A-Za-z0-9 < /dev/urandom | head -c 10; echo)"
-
-# BUILDKITE_AGENT_NAME format is {hostname}-{agent_idx}-{npu_card_num}cards
-agent_idx=$(echo "${BUILDKITE_AGENT_NAME}" | awk -F'-' '{print $(NF-1)}')
-echo "agent_idx: ${agent_idx}"
-builder_name="cachebuilder${agent_idx}"
-builder_cache_dir="/mnt/docker-cache${agent_idx}"
-mkdir -p ${builder_cache_dir}
-
-# Try building the docker image
-cat <<EOF | DOCKER_BUILDKIT=1 docker build \
-    --add-host pypi-cache:${PYPI_CACHE_HOST} \
-    --builder ${builder_name} --cache-from type=local,src=${builder_cache_dir} \
-                           --cache-to type=local,dest=${builder_cache_dir},mode=max \
-    --build-arg BUILDKITE_PULL_REQUEST="${BUILDKITE_PULL_REQUEST}" \
-    --build-arg BUILDKITE_PULL_REQUEST_REPO="${BUILDKITE_PULL_REQUEST_REPO}" \
-    --progress=plain --load -t ${image_name} -f - .
-FROM ${BASE_IMAGE_NAME}
-
-# Define environments
-ENV DEBIAN_FRONTEND=noninteractive
-
-RUN pip config set global.index-url http://pypi-cache:${PYPI_CACHE_PORT}/pypi/simple && \
-    pip config set global.trusted-host pypi-cache && \
-    apt-get update -y && \
-    apt-get install -y python3-pip git vim wget net-tools gcc g++ cmake libnuma-dev && \
-    rm -rf /var/cache/apt/* && \
-    rm -rf /var/lib/apt/lists/*
-
-# Install for pytest to make the docker build cache layer always valid
-RUN --mount=type=cache,target=/root/.cache/pip \
-    pip install pytest>=6.0  pytest-cov modelscope
-
-COPY . .
-
-# Install vllm-omni
-WORKDIR /workspace
-ARG VLLM_OMNI_REPO=https://github.com/vllm-project/vllm-omni.git
-ARG VLLM_OMNI_TAG=main
-ARG BUILDKITE_PULL_REQUEST
-ARG BUILDKITE_PULL_REQUEST_REPO
-RUN git config --global url."https://gh-proxy.test.osinfra.cn/https://github.com/".insteadOf "https://github.com/" && \
-    if [ "\$BUILDKITE_PULL_REQUEST" != "false" ] && [ -n "\$BUILDKITE_PULL_REQUEST" ]; then \
-        echo "Cloning and checking out PR #\$BUILDKITE_PULL_REQUEST..." && \
-        git clone \$VLLM_OMNI_REPO /workspace/vllm-omni && \
-        cd /workspace/vllm-omni && \
-        git fetch origin pull/\$BUILDKITE_PULL_REQUEST/head:pr-\$BUILDKITE_PULL_REQUEST && \
-        git checkout pr-\$BUILDKITE_PULL_REQUEST; \
-    else \
-        echo "Not a PR build, using main branch" && \
-        git clone --depth 1 \$VLLM_OMNI_REPO /workspace/vllm-omni; \
-    fi
-
-RUN --mount=type=cache,target=/root/.cache/pip \
-    export PIP_EXTRA_INDEX_URL=https://mirrors.huaweicloud.com/ascend/repos/pypi && \
-    source /usr/local/Ascend/ascend-toolkit/set_env.sh && \
-    source /usr/local/Ascend/nnal/atb/set_env.sh && \
-    export LD_LIBRARY_PATH=\$LD_LIBRARY_PATH:/usr/local/Ascend/ascend-toolkit/latest/`uname -i`-linux/devlib && \
-    python3 -m pip install -v -e /workspace/vllm-omni/
-
-ENV VLLM_WORKER_MULTIPROC_METHOD=spawn
-
-WORKDIR /workspace/vllm-omni
-CMD ["/bin/bash"]
-
-EOF
-
-# Setup cleanup
-remove_docker_container() {
-  docker rm -f "${container_name}" || true;
-  docker image rm -f "${image_name}" || true;
-  docker system prune -f || true;
-}
-trap remove_docker_container EXIT
-
-# Generate corresponding --device args based on BUILDKITE_AGENT_NAME
-# Ascend NPU BUILDKITE_AGENT_NAME format is {hostname}-{agent_idx}-{npu_card_num}cards, and agent_idx starts from 1.
-#   e.g. atlas-a2-001-1-2cards means this is the 1-th agent on atlas-a2-001 host, and it has 2 NPU cards.
-#   returns --device /dev/davinci0 --device /dev/davinci1
-parse_and_gen_devices() {
-    local input="$1"
-    local index cards_num
-    if [[ "$input" =~ ([0-9]+)-([0-9]+)cards$ ]]; then
-        index="${BASH_REMATCH[1]}"
-        cards_num="${BASH_REMATCH[2]}"
-    else
-        echo "parse error" >&2
-        return 1
-    fi
-
-    local devices=""
-    local i=0
-    while (( i < cards_num )); do
-        local dev_idx=$(((index - 1)*cards_num + i ))
-        devices="$devices --device /dev/davinci${dev_idx}"
-        ((i++))
-    done
-
-    # trim leading space
-    devices="${devices#"${devices%%[![:space:]]*}"}"
-    # Output devices: assigned to the caller variable
-    printf '%s' "$devices"
-}
-
-devices=$(parse_and_gen_devices "${BUILDKITE_AGENT_NAME}") || exit 1
-
-# Run the image and execute the Out-Of-Tree (OOT) platform interface test case on Ascend NPU hardware.
-# This test checks whether the OOT platform interface is functioning properly in conjunction with
-# the hardware plugin vllm-ascend.
-hf_model_cache_dir=/mnt/hf_cache${agent_idx}
-ms_model_cache_dir=/mnt/modelscope${agent_idx}
-mkdir -p ${hf_model_cache_dir}
-mkdir -p ${ms_model_cache_dir}
-docker run \
-    --init \
-    ${devices} \
-    --device /dev/davinci_manager \
-    --device /dev/devmm_svm \
-    --device /dev/hisi_hdc \
-    -v /usr/local/dcmi:/usr/local/dcmi \
-    -v /usr/local/bin/npu-smi:/usr/local/bin/npu-smi \
-    -v /usr/local/Ascend/driver/lib64/:/usr/local/Ascend/driver/lib64/ \
-    -v /usr/local/Ascend/driver/version.info:/usr/local/Ascend/driver/version.info \
-    -v /etc/ascend_install.info:/etc/ascend_install.info \
-    -v ${hf_model_cache_dir}:/root/.cache/huggingface \
-    -v ${ms_model_cache_dir}:/root/.cache/modelscope \
-    --network host \
-    --entrypoint="" \
-    --name "${container_name}" \
-    "${image_name}" \
-    bash -c '
-    set -e
-    VLLM_USE_MODELSCOPE=True pytest -s -v tests/e2e/offline_inference/test_qwen2_5_omni_expansion.py
-'

.buildkite/scripts/upload_pipeline.py

@@ -1,547 +0,0 @@
-#!/usr/bin/env python3
-# SPDX-License-Identifier: Apache-2.0
-"""Render and optionally upload Buildkite pipeline YAML with diff-aware logic.
-
-Bootstrap mode (pipeline.yml with __IMAGE_BUILD_IF__ placeholders):
-  - Detect docs-only, pytest skip-mark-only, or combined docs+skip-mark PR/main changes and
-    substitute skip-ci ``if`` expressions.
-
-Test pipeline mode (e.g. test-merge.yml):
-  - Drop steps or groups whose ``source_file_dependencies`` do not match changed files.
-  - ``source_file_dependencies`` may be set on a leaf step or on a group step.
-  - Always strip ``source_file_dependencies`` from uploaded YAML (Buildkite does not
-    recognize this uploader-only key).
-
-Usage:
-  python3 upload_pipeline.py [--upload] [--all | --e2e] <pipeline.yml>
-
-  # Bootstrap (replaces upload_pipeline_with_skip_ci.sh):
-  python3 upload_pipeline.py --upload .buildkite/pipeline.yml
-
-  # Test pipeline (replaces upload_test_pipeline_with_diff_skip.py):
-  python3 upload_pipeline.py --upload .buildkite/test-merge.yml
-
-  # Full suite (rebase pipeline): keep all steps, still strip the uploader-only key:
-  python3 upload_pipeline.py --upload --all .buildkite/test-merge.yml
-
-  # Nightly L2 E2E only: keep the E2E Test group, ignore source_file_dependencies:
-  python3 upload_pipeline.py --upload --e2e .buildkite/test-ready.yml
-"""
-
-from __future__ import annotations
-
-import argparse
-import os
-import re
-import subprocess
-import sys
-from pathlib import Path
-from typing import Any
-
-try:
-    import yaml
-except ImportError:
-    subprocess.check_call(
-        [sys.executable, "-m", "pip", "install", "-q", "pyyaml"],
-    )
-    import yaml
-
-LOG = "upload_pipeline"
-ROOT = Path(__file__).resolve().parent.parent.parent
-DOC_SEP = "\n---\n"
-BOOTSTRAP_MARKER = "__IMAGE_BUILD_IF__"
-E2E_GROUP_MARKER = "E2E Test"
-
-
-def _log(message: str) -> None:
-    print(f"{LOG}: {message}", file=sys.stderr)
-
-
-def _git(*args: str) -> subprocess.CompletedProcess[str]:
-    return subprocess.run(
-        ["git", *args],
-        check=False,
-        capture_output=True,
-        text=True,
-    )
-
-
-def resolve_diff_range() -> str | None:
-    """Return a git diff range for PR or main builds, or None when unavailable."""
-    is_pr = os.environ.get("BUILDKITE_PULL_REQUEST", "false") != "false" and os.environ.get(
-        "BUILDKITE_PULL_REQUEST", ""
-    )
-    commit = os.environ.get("BUILDKITE_COMMIT", "")
-
-    if is_pr:
-        base_branch = os.environ.get("BUILDKITE_PULL_REQUEST_BASE_BRANCH", "main")
-        base_ref = f"origin/{base_branch}"
-        if _git("rev-parse", "--verify", base_ref).returncode != 0:
-            _log(f"origin/{base_branch} not found locally; trying fetch")
-            _git("fetch", "--depth=200", "origin", base_branch)
-        if _git("rev-parse", "--verify", base_ref).returncode != 0:
-            if _git("rev-parse", "--verify", base_branch).returncode == 0:
-                base_ref = base_branch
-            else:
-                _log(f"cannot resolve PR base {base_branch}; using safe defaults")
-                return None
-        return f"{base_ref}...{commit}"
-
-    if os.environ.get("BUILDKITE_BRANCH", "") == "main":
-        if _git("rev-parse", "--verify", f"{commit}^").returncode != 0:
-            _log("main commit has no parent; using safe defaults")
-            return None
-        return f"{commit}^..{commit}"
-
-    _log("not PR/main build; using safe defaults")
-    return None
-
-
-def resolve_changed_files() -> list[str] | None:
-    """Return changed file paths, or None when diff cannot be resolved."""
-    diff_range = resolve_diff_range()
-    if diff_range is None:
-        return None
-
-    result = _git("diff", "--name-only", diff_range)
-    if result.returncode != 0:
-        _log(f"git diff failed ({diff_range}); using safe defaults")
-        return None
-
-    files = [line for line in result.stdout.splitlines() if line.strip()]
-    _log(f"{len(files)} changed file(s)")
-    return files
-
-
-def _is_doc_file(file_path: str) -> bool:
-    if not file_path:
-        return False
-    if file_path.startswith("docs/"):
-        return True
-    if file_path.endswith(".md"):
-        return True
-    return file_path == "mkdocs.yml"
-
-
-def is_docs_only_change(changed_files: list[str]) -> bool:
-    has_any = False
-    for file_path in changed_files:
-        if not file_path:
-            continue
-        has_any = True
-        if not _is_doc_file(file_path):
-            return False
-    return has_any
-
-
-def _is_test_python_file(file_path: str) -> bool:
-    path = Path(file_path)
-    return path.suffix == ".py" and bool(path.parts) and path.parts[0] == "tests"
-
-
-_SKIP_MARK_RE = re.compile(r"pytest\.mark\.skip(?:if)?\b|pytest\.skip\s*\(")
-_PYTESTMARK_SKIP_RE = re.compile(r"pytest\.mark\.skip\b")
-_PYTEST_MARK_ONLY_RE = re.compile(r"pytest\.mark\.\w+")
-
-
-def _paren_balance(line: str) -> int:
-    return line.count("(") - line.count(")")
-
-
-def _is_skip_mark_related_content(line: str) -> bool:
-    stripped = line.strip()
-    if not stripped:
-        return True
-    if stripped.startswith("#"):
-        return True
-    return _SKIP_MARK_RE.search(stripped) is not None
-
-
-def _is_pytestmark_adjacent_content(line: str) -> bool:
-    """Allow pytestmark list refactors that only add skip/skipif alongside existing marks."""
-    stripped = line.strip().rstrip(",")
-    if not stripped:
-        return True
-    if stripped in {"[", "]"}:
-        return True
-    if stripped.startswith("#"):
-        return True
-    if stripped.startswith("pytestmark"):
-        return True
-    return _PYTEST_MARK_ONLY_RE.search(stripped) is not None
-
-
-def diff_only_contains_skip_mark_changes(diff_text: str) -> bool:
-    """True when diff only edits skip marks or reformats pytestmark to add them."""
-    pending_depth = 0
-    saw_change = False
-    has_skip_mark_edit = False
-    for raw_line in diff_text.splitlines():
-        if raw_line.startswith("@@"):
-            pending_depth = 0
-            continue
-        if not (raw_line.startswith("+") or raw_line.startswith("-")):
-            continue
-        if raw_line.startswith("+++") or raw_line.startswith("---"):
-            continue
-
-        saw_change = True
-        content = raw_line[1:]
-
-        if pending_depth > 0:
-            pending_depth += _paren_balance(content)
-            if pending_depth < 0:
-                pending_depth = 0
-            continue
-
-        if _is_skip_mark_related_content(content):
-            has_skip_mark_edit = True
-            pending_depth = max(0, _paren_balance(content))
-            continue
-
-        if _is_pytestmark_adjacent_content(content):
-            continue
-
-        return False
-
-    return saw_change and has_skip_mark_edit
-
-
-def _git_file_diff(file_path: str, diff_range: str) -> str | None:
-    result = _git("diff", diff_range, "--", file_path)
-    if result.returncode != 0:
-        _log(f"skip-mark-only: git diff failed for {file_path}")
-        return None
-    return result.stdout
-
-
-def _is_new_file_diff(diff_text: str) -> bool:
-    return "new file mode" in diff_text or "--- /dev/null" in diff_text
-
-
-def _file_has_module_level_pytest_skip(content: str) -> bool:
-    """True when ``pytestmark`` applies unconditional ``pytest.mark.skip`` to the module."""
-    match = re.search(r"^pytestmark\s*=\s*\[(.*?)^\s*\]", content, re.MULTILINE | re.DOTALL)
-    if match is not None:
-        return _PYTESTMARK_SKIP_RE.search(match.group(1)) is not None
-    match = re.search(r"^pytestmark\s*=\s*pytest\.mark\.skip\b", content, re.MULTILINE)
-    return match is not None
-
-
-def _file_content_at_commit(file_path: str, commit: str) -> str | None:
-    result = _git("show", f"{commit}:{file_path}")
-    if result.returncode != 0:
-        _log(f"skip-mark-only: cannot read {file_path} at {commit}")
-        return None
-    return result.stdout
-
-
-def _qualifies_as_skip_mark_test_change(file_path: str, diff_range: str) -> bool:
-    diff_text = _git_file_diff(file_path, diff_range)
-    if diff_text is None:
-        return False
-    if not diff_text.strip():
-        _log(f"skip-mark-only: empty diff for {file_path}")
-        return False
-
-    if diff_only_contains_skip_mark_changes(diff_text):
-        return True
-
-    if not _is_new_file_diff(diff_text):
-        _log(f"skip-mark-only: non-skip changes in {file_path}")
-        return False
-
-    commit = os.environ.get("BUILDKITE_COMMIT", "")
-    if not commit:
-        return False
-    content = _file_content_at_commit(file_path, commit)
-    if content is None:
-        return False
-    if _file_has_module_level_pytest_skip(content):
-        _log(f"skip-mark-only: new test file with module-level pytest.mark.skip: {file_path}")
-        return True
-
-    _log(f"skip-mark-only: new test file without module-level pytest.mark.skip: {file_path}")
-    return False
-
-
-def is_skip_mark_only_change(changed_files: list[str], *, diff_range: str) -> bool:
-    """True when every changed file is a tests/ module with only skip-mark edits."""
-    has_any = False
-    for file_path in changed_files:
-        if not file_path:
-            continue
-        has_any = True
-        if not _is_test_python_file(file_path):
-            return False
-        if not _qualifies_as_skip_mark_test_change(file_path, diff_range):
-            return False
-    return has_any
-
-
-def is_docs_or_skip_mark_only_change(changed_files: list[str], *, diff_range: str) -> bool:
-    """True when each changed file is a doc path or a qualifying skip-mark test change."""
-    has_any = False
-    for file_path in changed_files:
-        if not file_path:
-            continue
-        has_any = True
-        if _is_doc_file(file_path):
-            continue
-        if _is_test_python_file(file_path) and _qualifies_as_skip_mark_test_change(file_path, diff_range):
-            continue
-        _log(f"docs/skip-mark-only: rejecting {file_path}")
-        return False
-    return has_any
-
-
-def resolve_skip_ci(changed_files: list[str] | None, *, diff_range: str | None = None) -> bool:
-    if changed_files is None:
-        _log("skip-ci=0 (could not resolve changed files)")
-        return False
-
-    if diff_range is not None and is_docs_or_skip_mark_only_change(changed_files, diff_range=diff_range):
-        if is_docs_only_change(changed_files):
-            _log("docs-only change detected; skip-ci=1")
-        elif is_skip_mark_only_change(changed_files, diff_range=diff_range):
-            _log("pytest skip-mark-only change detected; skip-ci=1")
-        else:
-            _log("docs + pytest skip-mark-only change detected; skip-ci=1")
-        return True
-
-    if diff_range is None and is_docs_only_change(changed_files):
-        _log("docs-only change detected; skip-ci=1")
-        return True
-
-    _log("non-doc/non-skip-mark changes detected; skip-ci=0")
-    return False
-
-
-def render_bootstrap_pipeline(text: str, *, skip_ci: bool) -> str:
-    if DOC_SEP in text:
-        _, continuation = text.split(DOC_SEP, 1)
-    else:
-        continuation = text
-
-    nightly_only = (
-        '(build.pull_request.labels includes "nightly-test") || (build.branch == "main" && build.env("NIGHTLY") == "1")'
-    )
-    nightly_main = 'build.branch == "main" && build.env("NIGHTLY") == "1"'
-    ready_pr = 'build.branch != "main" && build.pull_request.labels includes "ready"'
-    merge_main = 'build.branch == "main" && build.env("NIGHTLY") != "1" && build.env("WEEKLY") != "1"'
-    merge_pr = 'build.branch != "main" && build.pull_request.labels includes "merge-test"'
-    if skip_ci:
-        image_if = f"'{nightly_only}'"
-        ready_if = f"'({nightly_main})'"
-        merge_if = f"'({nightly_main})'"
-    else:
-        image_if = "'true'"
-        ready_if = f"'(({nightly_main}) || ({ready_pr}))'"
-        merge_if = f"'(({nightly_main}) || (({merge_main}) || ({merge_pr})))'"
-
-    return (
-        continuation.replace("__IMAGE_BUILD_IF__", image_if)
-        .replace("__UPLOAD_READY_IF__", ready_if)
-        .replace("__UPLOAD_MERGE_IF__", merge_if)
-    )
-
-
-def _matches_dependencies(changed_files: list[str], prefixes: list[str]) -> bool:
-    for path in changed_files:
-        for prefix in prefixes:
-            normalized = prefix.rstrip("/")
-            if path == normalized or path.startswith(f"{normalized}/"):
-                return True
-    return False
-
-
-def _strip_source_file_dependencies(step: dict[str, Any]) -> dict[str, Any]:
-    return {key: value for key, value in step.items() if key != "source_file_dependencies"}
-
-
-def _step_label(step: dict[str, Any]) -> str:
-    return str(step.get("group") or step.get("label") or "<step>")
-
-
-def _filter_steps(steps: list[Any], changed_files: list[str]) -> list[Any]:
-    filtered: list[Any] = []
-    for step in steps:
-        if not isinstance(step, dict):
-            filtered.append(step)
-            continue
-
-        deps = step.get("source_file_dependencies")
-        if deps is not None and not isinstance(deps, list):
-            raise ValueError(
-                f"source_file_dependencies must be a list in step {_step_label(step)!r}",
-            )
-        if deps is not None and not _matches_dependencies(changed_files, deps):
-            _log(
-                f"skip {_step_label(step)!r} (no changes under {deps})",
-            )
-            continue
-
-        nested = step.get("steps")
-        if nested is not None:
-            kept_nested = _filter_steps(nested, changed_files)
-            if not kept_nested:
-                _log(f"omit empty group {_step_label(step)!r}")
-                continue
-            new_step = _strip_source_file_dependencies(step)
-            new_step["steps"] = kept_nested
-            filtered.append(new_step)
-            continue
-
-        if deps is not None:
-            filtered.append(_strip_source_file_dependencies(step))
-        else:
-            filtered.append(step)
-
-    return filtered
-
-
-def _is_e2e_group(step: dict[str, Any]) -> bool:
-    group = step.get("group")
-    return isinstance(group, str) and E2E_GROUP_MARKER in group
-
-
-def _select_e2e_group_steps(steps: list[Any]) -> list[Any]:
-    selected = [step for step in steps if isinstance(step, dict) and _is_e2e_group(step)]
-    if not selected:
-        _log(f"no group matching {E2E_GROUP_MARKER!r} found")
-    else:
-        _log(f"keep {len(selected)} group(s) matching {E2E_GROUP_MARKER!r}")
-    return selected
-
-
-def _strip_uploader_metadata_from_steps(steps: list[Any]) -> list[Any]:
-    """Remove uploader-only keys while keeping all steps (no diff filtering)."""
-    stripped: list[Any] = []
-    for step in steps:
-        if not isinstance(step, dict):
-            stripped.append(step)
-            continue
-
-        deps = step.get("source_file_dependencies")
-        if deps is not None and not isinstance(deps, list):
-            raise ValueError(
-                f"source_file_dependencies must be a list in step {_step_label(step)!r}",
-            )
-
-        nested = step.get("steps")
-        new_step = _strip_source_file_dependencies(step)
-        if nested is not None:
-            new_step["steps"] = _strip_uploader_metadata_from_steps(nested)
-        stripped.append(new_step)
-
-    return stripped
-
-
-def render_test_pipeline(
-    doc: dict[str, Any],
-    changed_files: list[str] | None,
-    *,
-    e2e_only: bool = False,
-) -> dict[str, Any]:
-    steps = doc.get("steps")
-    if not isinstance(steps, list):
-        return doc
-    if e2e_only:
-        steps = _select_e2e_group_steps(steps)
-        steps = _strip_uploader_metadata_from_steps(steps)
-    elif changed_files is not None:
-        steps = _filter_steps(steps, changed_files)
-    else:
-        steps = _strip_uploader_metadata_from_steps(steps)
-    return {**doc, "steps": steps}
-
-
-def resolve_pipeline_path(arg: str) -> Path:
-    path = Path(arg)
-    if path.is_absolute():
-        return path
-    return ROOT / path
-
-
-def render_pipeline(
-    path: Path,
-    *,
-    force_all: bool = False,
-    e2e_only: bool = False,
-) -> str:
-    text = path.read_text(encoding="utf-8")
-    diff_range = resolve_diff_range()
-    changed_files = resolve_changed_files()
-
-    # ``--all`` forces the keep-all-steps path (no diff-aware skipping) while still
-    # stripping ``source_file_dependencies``. Used by the rebase pipeline so main builds
-    # run the full e2e suite (see .buildkite/rebase-pipeline.yaml).
-    if force_all or e2e_only:
-        changed_files = None
-
-    if BOOTSTRAP_MARKER in text:
-        rendered = render_bootstrap_pipeline(
-            text,
-            skip_ci=resolve_skip_ci(changed_files, diff_range=diff_range),
-        )
-        return rendered
-
-    doc = yaml.safe_load(text)
-    if not isinstance(doc, dict):
-        raise ValueError(f"invalid pipeline YAML: {path}")
-
-    doc = render_test_pipeline(doc, changed_files, e2e_only=e2e_only)
-
-    return yaml.safe_dump(doc, sort_keys=False)
-
-
-def upload_to_buildkite(content: str) -> None:
-    subprocess.run(
-        ["buildkite-agent", "pipeline", "upload"],
-        input=content,
-        text=True,
-        check=True,
-    )
-
-
-def main() -> int:
-    parser = argparse.ArgumentParser(description=__doc__)
-    parser.add_argument(
-        "pipeline",
-        nargs="?",
-        default=".buildkite/pipeline.yml",
-        help="Pipeline YAML path (default: .buildkite/pipeline.yml)",
-    )
-    parser.add_argument(
-        "--upload",
-        action="store_true",
-        help="Pipe rendered YAML to buildkite-agent pipeline upload",
-    )
-    mode = parser.add_mutually_exclusive_group()
-    mode.add_argument(
-        "--all",
-        action="store_true",
-        help="Keep all steps (disable diff-aware skipping); still strips source_file_dependencies",
-    )
-    mode.add_argument(
-        "--e2e",
-        action="store_true",
-        help="Keep only the E2E Test group; disable diff-aware skipping for those steps",
-    )
-    args = parser.parse_args()
-
-    path = resolve_pipeline_path(args.pipeline)
-    if not path.is_file():
-        _log(f"missing pipeline file: {path}")
-        return 1
-
-    rendered = render_pipeline(path, force_all=args.all, e2e_only=args.e2e)
-    if args.upload:
-        upload_to_buildkite(rendered)
-    else:
-        sys.stdout.write(rendered)
-    return 0
-
-
-if __name__ == "__main__":
-    raise SystemExit(main())

.buildkite/test-amd.yaml

@@ -1,118 +0,0 @@
-steps:
-
-- label: "Diffusion Model Test"
-  timeout_in_minutes: 30
-  agent_pool: mi325_2
-  depends_on: amd-build
-  mirror_hardwares: [amdproduction]
-  grade: Blocking
-  commands:
-    - export GPU_ARCHS=gfx942
-    - pytest -s -v tests/e2e/offline_inference/test_z_image.py
-
-- label: "Diffusion Images API LoRA E2E"
-  timeout_in_minutes: 30
-  agent_pool: mi325_1
-  depends_on: amd-build
-  mirror_hardwares: [amdproduction]
-  grade: Blocking
-  commands:
-    - export GPU_ARCHS=gfx942
-    - export VLLM_LOGGING_LEVEL=DEBUG
-    - export VLLM_WORKER_MULTIPROC_METHOD=spawn
-    - pytest -s -v tests/diffusion/lora/test_images_generations_lora.py
-
-- label: "Diffusion Model CPU offloading Test"
-  timeout_in_minutes: 20
-  agent_pool: mi325_1
-  depends_on: amd-build
-  mirror_hardwares: [amdproduction]
-  grade: Blocking
-  commands:
-    - export GPU_ARCHS=gfx942
-    - export VLLM_LOGGING_LEVEL=DEBUG
-    - export VLLM_WORKER_MULTIPROC_METHOD=spawn
-    - pytest -s -v tests/diffusion/offloader/test_diffusion_cpu_offload.py
-    - pytest -s -v tests/diffusion/offloader/test_diffusion_layerwise_offload.py
-
-- label: "Diffusion Cache Backend Test"
-  timeout_in_minutes: 15
-  agent_pool: mi325_1
-  depends_on: amd-build
-  mirror_hardwares: [amdproduction]
-  grade: Blocking
-  commands:
-    - export GPU_ARCHS=gfx942
-    - export VLLM_LOGGING_LEVEL=DEBUG
-    - export VLLM_WORKER_MULTIPROC_METHOD=spawn
-    - pytest -s -v -m 'core_model and cache and diffusion and not distributed_rocm and MI325'
-
-- label: "Diffusion Sequence Parallelism Test"
-  timeout_in_minutes: 20
-  agent_pool: mi325_2
-  depends_on: amd-build
-  mirror_hardwares: [amdproduction]
-  grade: Blocking
-  commands:
-    - export GPU_ARCHS=gfx942
-    - export VLLM_LOGGING_LEVEL=DEBUG
-    - export VLLM_WORKER_MULTIPROC_METHOD=spawn
-    - pytest -s -v tests/diffusion/distributed/test_sequence_parallel.py -m core_model
-
-- label: "Diffusion Tensor Parallelism Test"
-  timeout_in_minutes: 20
-  agent_pool: mi325_2
-  depends_on: amd-build
-  commands:
-    - export GPU_ARCHS=gfx942
-    - export VLLM_LOGGING_LEVEL=DEBUG
-    - export VLLM_WORKER_MULTIPROC_METHOD=spawn
-    - pytest -s -v tests/diffusion/distributed/test_zimage_parallelism.py
-
-- label: "Diffusion GPU Worker Test"
-  timeout_in_minutes: 20
-  agent_pool: mi325_2
-  depends_on: amd-build
-  mirror_hardwares: [amdproduction]
-  grade: Blocking
-  commands:
-    - pytest -s -v tests/diffusion/test_diffusion_worker.py
-
-- label: "Omni Model Test Qwen2-5-Omni"
-  timeout_in_minutes: 15
-  agent_pool: mi325_2
-  depends_on: amd-build
-  mirror_hardwares: [amdproduction]
-  grade: Blocking
-  commands:
-    - export GPU_ARCHS=gfx942
-    - export VLLM_LOGGING_LEVEL=DEBUG
-    - export VLLM_WORKER_MULTIPROC_METHOD=spawn
-    - pytest -s -v tests/e2e/offline_inference/test_qwen2_5_omni_expansion.py
-    - pytest -s -v tests/engine/test_async_omni_engine_abort.py
-
-- label: "Omni Model Test Qwen3-Omni"
-  timeout_in_minutes: 15
-  agent_pool: mi325_2
-  depends_on: amd-build
-  mirror_hardwares: [amdproduction]
-  grade: Blocking
-  commands:
-    - export VLLM_LOGGING_LEVEL=DEBUG
-    - export VLLM_WORKER_MULTIPROC_METHOD=spawn
-    - export VLLM_ROCM_USE_AITER=0
-    - pytest -s -v tests/e2e/offline_inference/test_qwen3_omni.py
-    - pytest -s -v tests/e2e/online_serving/test_qwen3_omni.py
-
-
-- label: "Omni Sleep Mode Test"
-  timeout_in_minutes: 40
-  agent_pool: mi325_2
-  depends_on: amd-build
-  mirror_hardwares: [amdproduction]
-  grade: Blocking
-  commands:
-    - export GPU_ARCHS=gfx942
-    - export VLLM_LOGGING_LEVEL=DEBUG
-    - export VLLM_WORKER_MULTIPROC_METHOD=spawn
-    - pytest -s -v tests/entrypoints/test_omni_sleep_mode.py -m "advanced_model and omni and MI325" --run-level "advanced_model"

.buildkite/test-merge.yml

@@ -1,645 +0,0 @@
-env:
-  VLLM_WORKER_MULTIPROC_METHOD: spawn
-  VLLM_USE_FLASHINFER_MOE_FP16: 0
-  HF_HUB_DOWNLOAD_TIMEOUT: 300
-  HF_HUB_ETAG_TIMEOUT: 60
-
-steps:
-  - group: ":card_index_dividers: Simple Test"
-    depends_on: upload-merge-pipeline
-    steps:
-      - label: "Simple · Diffusion & Model Executor Test"
-        commands:
-          - "timeout 20m pytest -sv tests/diffusion tests/model_executor -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Simple · Other Test"
-        commands:
-          - "timeout 20m pytest -sv -m 'core_model and cpu' --ignore=tests/diffusion --ignore=tests/model_executor --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-  - group: ":card_index_dividers: Diffusion Test"
-    depends_on: upload-merge-pipeline
-    steps:
-      - label: "Diffusion · LoRA Test"
-        timeout_in_minutes: 30
-        commands:
-          - pytest -sv tests/diffusion/lora/ -m 'advanced_model and cuda' --run-level "advanced_model"
-        agents:
-          queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-  - label: "Distributed Test"
-    depends_on: upload-merge-pipeline
-    timeout_in_minutes: 60
-    commands:
-      - |
-        timeout 55m bash -c "
-          set -e
-          export VLLM_IMAGE_FETCH_TIMEOUT=60
-          pytest -s -v tests/distributed -m 'advanced_model and cuda and H100' --run-level 'advanced_model'
-        "
-    agents:
-      queue: "mithril-h100-pool"
-    plugins:
-      - kubernetes:
-          podSpec:
-            containers:
-              - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                resources:
-                  limits:
-                    nvidia.com/gpu: 1
-                volumeMounts:
-                  - name: devshm
-                    mountPath: /dev/shm
-                  - name: hf-cache
-                    mountPath: /root/.cache/huggingface
-                env:
-                  - name: HF_HOME
-                    value: /root/.cache/huggingface
-                  - name: HF_TOKEN
-                    valueFrom:
-                      secretKeyRef:
-                        name: hf-token-secret
-                        key: token
-            nodeSelector:
-              node.kubernetes.io/instance-type: gpu-h100-sxm
-            volumes:
-              - name: devshm
-                emptyDir:
-                  medium: Memory
-              - name: hf-cache
-                hostPath:
-                  path: /mnt/hf-cache
-                  type: DirectoryOrCreate
-
-  - group: ":card_index_dividers: Entrypoints Test"
-    depends_on: upload-merge-pipeline
-    steps:
-      - label: "Entrypoint Test with H100"
-        timeout_in_minutes: 60
-        commands:
-          - pytest -s -v tests/entrypoints/ -m "advanced_model and cuda and H100" --run-level "advanced_model"
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 2
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "Entrypoint Test with L4"
-        timeout_in_minutes: 30
-        commands:
-          - pytest -s -v tests/entrypoints/ -m "advanced_model and cuda and L4" --run-level "advanced_model"
-        agents:
-          queue: "gpu_4_queue" # g6.12xlarge instance on AWS, has 4 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-  - group: ":card_index_dividers: E2E Test"
-    depends_on: upload-merge-pipeline
-    steps:
-      - label: "TTS · Qwen3-TTS CustomVoice Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen3_tts_customvoice.py
-          - tests/e2e/offline_inference/test_qwen3_tts_customvoice.py
-          - vllm_omni/model_executor/models/qwen3_tts/
-          - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
-          - vllm_omni/deploy/qwen3_tts.yaml
-        commands:
-          - |
-            timeout 40m bash -c "
-              export VLLM_LOGGING_LEVEL=DEBUG
-              export VLLM_ALLOW_LONG_MAX_MODEL_LEN=1
-              pytest -s -v tests/e2e/online_serving/test_qwen3_tts_customvoice.py tests/e2e/offline_inference/test_qwen3_tts_customvoice.py -m 'advanced_model and cuda' --run-level 'advanced_model'
-            "
-        agents:
-          queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "TTS · Qwen3-TTS Base Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen3_tts_base.py
-          - tests/e2e/offline_inference/test_qwen3_tts_base.py
-          - vllm_omni/model_executor/models/qwen3_tts/
-          - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
-          - vllm_omni/deploy/qwen3_tts.yaml
-        commands:
-          - |
-            timeout 40m bash -c "
-              export VLLM_LOGGING_LEVEL=DEBUG
-              export VLLM_ALLOW_LONG_MAX_MODEL_LEN=1
-              pytest -s -v tests/e2e/online_serving/test_qwen3_tts_base.py tests/e2e/offline_inference/test_qwen3_tts_base.py -m 'advanced_model and cuda' --run-level 'advanced_model'
-            "
-        agents:
-          queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "TTS · Higgs-Audio-V3 Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_higgs_audio_v3.py
-          - vllm_omni/model_executor/models/higgs_audio_v3/
-          - vllm_omni/model_executor/stage_input_processors/higgs_audio_v3.py
-          - vllm_omni/deploy/higgs_multimodal_qwen3.yaml
-        commands:
-          - |
-            timeout 20m bash -c "
-              export VLLM_WORKER_MULTIPROC_METHOD=spawn
-              export VLLM_USE_DEEP_GEMM=0
-              export VLLM_MOE_USE_DEEP_GEMM=0
-              pytest -s -v tests/e2e/online_serving/test_higgs_audio_v3.py -m 'advanced_model' --run-level 'advanced_model'
-            "
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "TTS · IndexTTS2 Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_indextts2.py
-          - tests/e2e/offline_inference/test_indextts2.py
-          - tests/model_executor/models/indextts2/
-          - tests/model_executor/stage_input_processors/test_indextts2.py
-          - tests/assets/indextts2/
-          - vllm_omni/model_executor/models/indextts2/
-          - vllm_omni/model_executor/stage_input_processors/indextts2.py
-          - vllm_omni/deploy/indextts2.yaml
-          - vllm_omni/deploy/indextts2_low_latency.yaml
-          - vllm_omni/entrypoints/openai/serving_speech.py
-          - vllm_omni/entrypoints/openai/tts_adapters/indextts2.py
-          - vllm_omni/config/pipeline_registry.py
-          - vllm_omni/model_executor/models/registry.py
-        timeout_in_minutes: 60
-        commands:
-          - |
-            timeout 45m bash -c "
-              export VLLM_LOGGING_LEVEL=DEBUG
-              pytest -s -v tests/e2e/online_serving/test_indextts2.py tests/e2e/offline_inference/test_indextts2.py -m 'advanced_model and cuda' --run-level 'advanced_model'
-            "
-        agents:
-          queue: "gpu_4_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              gpus: all
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Omni · Qwen3-Omni Test"
-        source_file_dependencies:
-          - tests/e2e/offline_inference/test_qwen3_omni.py
-          - tests/e2e/online_serving/test_qwen3_omni.py
-          - vllm_omni/model_executor/models/qwen3_omni/
-          - vllm_omni/model_executor/stage_input_processors/qwen3_omni.py
-          - vllm_omni/deploy/qwen3_omni_moe.yaml
-        timeout_in_minutes: 50
-        commands:
-          - pytest -s -v tests/e2e/offline_inference/test_qwen3_omni.py tests/e2e/online_serving/test_qwen3_omni.py -m 'advanced_model and cuda' --run-level 'advanced_model'
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 2
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "TTS · VoxCPM2 Test"
-        source_file_dependencies:
-          - tests/e2e/offline_inference/test_voxcpm2_tts.py
-          - tests/e2e/online_serving/test_voxcpm2_tts.py
-          - vllm_omni/model_executor/models/voxcpm2/
-          - vllm_omni/deploy/voxcpm2.yaml
-        timeout_in_minutes: 30
-        commands:
-          - |
-            timeout 30m bash -c "
-              export VLLM_LOGGING_LEVEL=DEBUG
-              export VLLM_WORKER_MULTIPROC_METHOD=spawn
-              pytest -s -v tests/e2e/offline_inference/test_voxcpm2_tts.py tests/e2e/online_serving/test_voxcpm2_tts.py -m 'advanced_model and cuda' --run-level 'advanced_model'
-            "
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "TTS · SoulX-Singer Test"
-        source_file_dependencies:
-          - tests/e2e/offline_inference/test_soulxsinger.py
-          - vllm_omni/diffusion/models/soulx_singer/
-          - vllm_omni/deploy/soulxsinger_svc.yaml
-          - vllm_omni/deploy/soulxsinger_svs.yaml
-        timeout_in_minutes: 30
-        commands:
-          - |
-            timeout 30m bash -c "
-              export VLLM_LOGGING_LEVEL=DEBUG
-              export VLLM_WORKER_MULTIPROC_METHOD=spawn
-              pytest -s -v tests/e2e/offline_inference/test_soulxsinger.py -k 'svc or precomputed' -m 'advanced_model and cuda' --run-level 'advanced_model'
-            "
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Diffusion · Qwen Image Edit Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen_image_edit.py
-          - vllm_omni/diffusion/models/qwen_image/
-        timeout_in_minutes: 20
-        commands:
-          - pytest -s -v tests/e2e/online_serving/test_qwen_image_edit.py -m 'advanced_model and cuda' --run-level 'advanced_model'
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "Diffusion · Bagel Test"
-        source_file_dependencies:
-          - tests/e2e/offline_inference/test_bagel.py
-          - tests/e2e/online_serving/test_bagel.py
-          - vllm_omni/model_executor/models/bagel/
-          - vllm_omni/diffusion/models/bagel/
-          - vllm_omni/model_executor/stage_input_processors/bagel.py
-          - vllm_omni/deploy/bagel.yaml
-        timeout_in_minutes: 20
-        commands:
-          - pytest -s -v tests/e2e/offline_inference/test_bagel.py tests/e2e/online_serving/test_bagel.py -m 'advanced_model and cuda' --run-level 'advanced_model'
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-
-      - label: "Diffusion · Qwen Image Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen_image.py
-          - vllm_omni/diffusion/models/qwen_image/
-        commands:
-          - |
-            timeout 40m bash -c '
-              export VLLM_IMAGE_FETCH_TIMEOUT=60
-              pytest -s -v tests/e2e/online_serving/test_qwen_image.py -m 'advanced_model and cuda' --run-level 'advanced_model'
-            '
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "Diffusion · Qwen Image Layered Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen_image_layered.py
-          - vllm_omni/diffusion/models/qwen_image/
-        commands:
-          - |
-            timeout 40m bash -c "
-              export VLLM_IMAGE_FETCH_TIMEOUT=60
-              pytest -s -v tests/e2e/online_serving/test_qwen_image_layered.py -m 'advanced_model and cuda' --run-level 'advanced_model'
-            "
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "Diffusion · Wan22 Test"
-        source_file_dependencies:
-          - tests/e2e/offline_inference/test_wan22_t2v.py
-          - tests/e2e/online_serving/test_wan22_t2v.py
-          - vllm_omni/diffusion/models/wan2_2/
-        commands:
-          - |
-            timeout 40m bash -c "
-              export VLLM_IMAGE_FETCH_TIMEOUT=60
-              pytest -s -v tests/e2e/offline_inference/test_wan22_t2v.py tests/e2e/online_serving/test_wan22_t2v.py -m 'advanced_model and cuda and diffusion' --run-level 'advanced_model'
-            "
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "Diffusion · Z Image Test"
-        source_file_dependencies:
-          - tests/e2e/offline_inference/test_z_image.py
-          - tests/e2e/online_serving/test_z_image.py
-          - vllm_omni/diffusion/models/z_image/
-        timeout_in_minutes: 30
-        commands:
-          - pytest -s -v tests/e2e/offline_inference/test_z_image.py tests/e2e/online_serving/test_z_image.py -m 'advanced_model and cuda and diffusion' --run-level 'advanced_model'
-        agents:
-          queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Diffusion · FLUX.2 Klein Test"
-        source_file_dependencies:
-          - tests/e2e/offline_inference/test_flux2_klein.py
-          - vllm_omni/diffusion/models/flux2_klein/
-        timeout_in_minutes: 30
-        commands:
-          - pytest -s -v tests/e2e/offline_inference/test_flux2_klein.py -m 'advanced_model and cuda and diffusion' --run-level 'advanced_model'
-        agents:
-          queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"

.buildkite/test-npu-nightly.yaml

@@ -1,52 +0,0 @@
-steps:
-  - group: ":card_index_dividers: Diffusion X2V Model Test"
-    key: nightly-diffusion-x2v-group
-    depends_on: upload-nightly-pipeline
-    if: >-
-      build.env("NIGHTLY") == "1" ||
-      build.pull_request.labels includes "nightly-test" ||
-      build.pull_request.labels includes "diffusion-x2v-test"
-    steps:
-      - label: ":full_moon: Diffusion X2V · Function Test"
-        agents:
-          queue: "ascend-a2b3"
-          resource_class: "npu-4"
-        image: "swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/vllm-omni-ci-npu:${BUILDKITE_COMMIT}"
-        plugins:
-          - kubernetes:
-              podSpecPatch:
-                imagePullSecrets:
-                  - name: swr-secret
-        env:
-          VLLM_WORKER_MULTIPROC_METHOD: spawn
-          HF_TOKEN: "${HF_TOKEN}"
-        commands:
-          - pytest -s -v tests/e2e/online_serving/test_wan22_expansion.py -k npu_i2v --run-level "full_model"
-  - group: ":card_index_dividers: TTS Model Test"
-    key: nightly-tts-test-group
-    depends_on: upload-nightly-pipeline
-    if: build.env("NIGHTLY") == "1" || build.pull_request.labels includes "nightly-test" || build.pull_request.labels includes "tts-test"
-    steps:
-      - label: ":full_moon: TTS · Perf Test"
-        agents:
-          queue: "ascend-a3"
-          resource_class: "npu-2"
-        image: "swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/vllm-omni-ci-npu:${BUILDKITE_COMMIT}-a3"
-        plugins:
-          - kubernetes:
-              podSpecPatch:
-                imagePullSecrets:
-                  - name: swr-secret
-        env:
-          VLLM_WORKER_MULTIPROC_METHOD: spawn
-          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
-          HF_TOKEN: "${HF_TOKEN}"
-        commands:
-          - export BENCHMARK_DIR=tests/dfx/perf/results
-          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
-          - |
-            set +e
-            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py -m npu --test-config-file tests/dfx/perf/tests/test_tts.json
-            EXIT=$$?
-            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
-            exit $$EXIT

.buildkite/test-npu-ready.yaml

@@ -1,21 +0,0 @@
-steps:
-  - group: ":card_index_dividers: TTS Model Test"
-    key: ready-tts-group
-    depends_on: upload-ready-pipeline
-    steps:
-      - label: ":sound: TTS · Qwen3-TTS CustomVoice Test"
-        agents:
-          queue: "ascend-a2b3"
-          resource_class: "npu-1"
-        image: "swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/vllm-omni-ci-npu:${BUILDKITE_COMMIT}"
-        plugins:
-          - kubernetes:
-              podSpecPatch:
-                imagePullSecrets:
-                  - name: swr-secret
-        env:
-          VLLM_WORKER_MULTIPROC_METHOD: spawn
-          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
-          HF_TOKEN: "${HF_TOKEN}"
-        commands:
-          - pytest -s -v tests/e2e/online_serving/test_qwen3_tts_customvoice.py -m 'core_model' --run-level 'core_model'

.buildkite/test-ready.yml

@@ -1,707 +0,0 @@
-env:
-  VLLM_WORKER_MULTIPROC_METHOD: spawn
-  VLLM_USE_FLASHINFER_MOE_FP16: 0
-  HF_HUB_DOWNLOAD_TIMEOUT: 300
-  HF_HUB_ETAG_TIMEOUT: 60
-
-steps:
-  - group: ":card_index_dividers: Simple Test"
-    depends_on: upload-ready-pipeline
-    steps:
-      - label: "Simple · Diffusion & Model Executor Test"
-        commands:
-          - "timeout 20m pytest -sv tests/diffusion tests/model_executor -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Simple · Other Test"
-        commands:
-          - "timeout 20m pytest -sv -m 'core_model and cpu' --ignore=tests/diffusion --ignore=tests/model_executor --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-  - label: "Custom Pipeline Test"
-    depends_on: upload-ready-pipeline
-    commands:
-      - timeout 20m pytest -s -v tests/e2e/offline_inference/custom_pipeline/ -m "core_model"
-    agents:
-      queue: "gpu_1_queue"
-    plugins:
-      - docker#v5.2.0:
-          image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-          always-pull: true
-          propagate-environment: true
-          environment:
-            - "HF_HOME=/fsx/hf_cache"
-          volumes:
-            - "/fsx/hf_cache:/fsx/hf_cache"
-
-  - group: ":card_index_dividers: Diffusion Test"
-    depends_on: upload-ready-pipeline
-    steps:
-      - label: "Diffusion · Offloader Test"
-        commands:
-          - timeout 40m pytest -s -v tests/diffusion/offloader -m "core_model and cuda" --run-level "core_model"
-        agents:
-          queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Diffusion · Cache Test"
-        commands:
-          - timeout 15m pytest -sv tests/diffusion/cache -m 'core_model and cuda' --run-level "core_model"
-        agents:
-          queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Diffusion · Batch Test"
-        commands:
-          - timeout 15m pytest -sv tests/diffusion/batching -m 'core_model and cuda' --run-level "core_model"
-        agents:
-          queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Diffusion · Distributed Test"
-        commands:
-          - timeout 20m pytest -sv tests/diffusion/distributed -m 'core_model and cuda' --run-level "core_model"
-        agents:
-          queue: "gpu_4_queue" # g6.12xlarge instance on AWS, has 4 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Diffusion · Model Test"
-        commands:
-          - timeout 15m pytest -sv tests/diffusion/models/ -m 'core_model and cuda' --run-level "core_model"
-        agents:
-          queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Diffusion · Other Test"
-        commands:
-          - timeout 15m pytest -sv tests/diffusion/*.py -m 'core_model and cuda' --run-level "core_model"
-        agents:
-          queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-
-  - group: ":card_index_dividers: Tiny Model Tests [Single GPU]"
-    depends_on: upload-ready-pipeline
-    steps:
-      - label: "Tiny Model Tests · Diffusion (base)"
-        commands:
-          - timeout 15m pytest -sv tests/model_tests/diffusion -m 'core_model and cuda' --run-level "core_model"
-        agents:
-          # Ready label only runs the cases that don't need parallelism for tiny
-          # model tests, which is usually just a smoke test for TTI and I2I
-          queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-  - label: "Engine&Model Executor Test"
-    depends_on: upload-ready-pipeline
-    commands:
-      - |
-        timeout 15m bash -c "
-                  pytest -sv tests/engine/ tests/model_executor/ -m 'core_model and cuda' --run-level 'core_model'
-        "
-    agents:
-      queue: "gpu_1_queue" # g6.12xlarge instance on AWS, has 4 L4 GPU
-    plugins:
-      - docker#v5.2.0:
-          image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-          always-pull: true
-          propagate-environment: true
-          shm-size: "8gb"
-          environment:
-            - "HF_HOME=/fsx/hf_cache"
-            - "HF_TOKEN"
-          volumes:
-            - "/fsx/hf_cache:/fsx/hf_cache"
-
-
-  - label: "Distributed Test"
-    depends_on: upload-ready-pipeline
-    commands:
-      - |
-        timeout 20m bash -c "
-          pytest -sv tests/distributed/ -m 'core_model and cuda and L4' --run-level 'core_model'
-        "
-    agents:
-      queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-    plugins:
-      - docker#v5.2.0:
-          image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-          always-pull: true
-          propagate-environment: true
-          shm-size: "8gb"
-          environment:
-            - "HF_HOME=/fsx/hf_cache"
-            - "HF_TOKEN"
-          volumes:
-            - "/fsx/hf_cache:/fsx/hf_cache"
-
-  - label: "Entrypoints Test"
-    depends_on: upload-ready-pipeline
-    commands:
-      - |
-        timeout 30m bash -c "
-          export VLLM_WORKER_MULTIPROC_METHOD=spawn
-          pytest -sv tests/entrypoints/ -m 'core_model and cuda' --run-level 'core_model'
-        "
-    agents:
-      queue: "gpu_1_queue"
-    plugins:
-      - docker#v5.2.0:
-          image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-          always-pull: true
-          propagate-environment: true
-          shm-size: "8gb"
-          environment:
-            - "HF_HOME=/fsx/hf_cache"
-            - "HF_TOKEN"
-          volumes:
-            - "/fsx/hf_cache:/fsx/hf_cache"
-
-  - group: ":card_index_dividers: E2E Test"
-    depends_on: upload-ready-pipeline
-    steps:
-      - label: "RLHF · VeRL-Omni E2E Test"
-        source_file_dependencies:
-          - tests/e2e/offline_inference/rlhf_test/test_verl_omni_e2e.py
-          - tests/e2e/offline_inference/custom_pipeline/
-          - vllm_omni/diffusion/
-          - vllm_omni/engine/
-          - vllm_omni/entrypoints/
-          - vllm_omni/inputs/
-          - vllm_omni/outputs/
-        commands:
-          - |
-            timeout 20m bash -c "
-              pytest -s -v tests/e2e/offline_inference/rlhf_test/test_verl_omni_e2e.py
-            "
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Omni · Qwen3-Omni Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen3_omni.py
-          - vllm_omni/model_executor/models/qwen3_omni/
-          - vllm_omni/model_executor/stage_input_processors/qwen3_omni.py
-          - vllm_omni/deploy/qwen3_omni_moe.yaml
-        commands:
-          - |
-            timeout 20m bash -c "
-              pytest -s -v tests/e2e/online_serving/test_qwen3_omni.py -m 'core_model' --run-level 'core_model'
-            "
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 2
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "TTS · VoxCPM2 Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_voxcpm2_tts.py
-          - vllm_omni/model_executor/models/voxcpm2/
-          - vllm_omni/deploy/voxcpm2.yaml
-        timeout_in_minutes: 30
-        commands:
-          - |
-            timeout 30m bash -c "
-              export VLLM_LOGGING_LEVEL=DEBUG
-              export VLLM_WORKER_MULTIPROC_METHOD=spawn
-              pytest -s -v tests/e2e/online_serving/test_voxcpm2_tts.py -m 'core_model' --run-level 'core_model'
-            "
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "TTS · Qwen3-TTS CustomVoice Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen3_tts_customvoice.py
-          - vllm_omni/model_executor/models/qwen3_tts/
-          - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
-          - vllm_omni/deploy/qwen3_tts.yaml
-        commands:
-          - |
-            timeout 20m bash -c "
-              export VLLM_ALLOW_LONG_MAX_MODEL_LEN=1
-              pytest -s -v tests/e2e/online_serving/test_qwen3_tts_customvoice.py -m 'core_model' --run-level 'core_model'
-            "
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "TTS · Qwen3-TTS Base Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen3_tts_base.py
-          - vllm_omni/model_executor/models/qwen3_tts/
-          - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
-          - vllm_omni/deploy/qwen3_tts.yaml
-        commands:
-          - |
-            timeout 20m bash -c "
-              export VLLM_WORKER_MULTIPROC_METHOD=spawn
-              export VLLM_ALLOW_LONG_MAX_MODEL_LEN=1
-              export VLLM_OMNI_USE_V2_RUNNER=1
-              pytest -s -v tests/e2e/online_serving/test_qwen3_tts_base.py -m 'core_model' --run-level 'core_model'
-            "
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "TTS · Higgs-Audio-V3 Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_higgs_audio_v3.py
-          - vllm_omni/model_executor/models/higgs_audio_v3/
-          - vllm_omni/model_executor/stage_input_processors/higgs_audio_v3.py
-          - vllm_omni/deploy/higgs_multimodal_qwen3.yaml
-        commands:
-          - |
-            timeout 20m bash -c "
-              export VLLM_WORKER_MULTIPROC_METHOD=spawn
-              export VLLM_USE_DEEP_GEMM=0
-              export VLLM_MOE_USE_DEEP_GEMM=0
-              pytest -s -v tests/e2e/online_serving/test_higgs_audio_v3.py -m 'core_model' --run-level 'core_model'
-            "
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Diffusion · Qwen Image Edit Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen_image_edit.py
-          - vllm_omni/diffusion/models/qwen_image/
-        commands:
-          - |
-            timeout 20m bash -c "
-              pytest -s -v tests/e2e/online_serving/test_qwen_image_edit.py -m 'core_model' --run-level 'core_model'
-            "
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "Diffusion · Bagel Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_bagel.py
-          - vllm_omni/model_executor/models/bagel/
-          - vllm_omni/diffusion/models/bagel/
-          - vllm_omni/model_executor/stage_input_processors/bagel.py
-          - vllm_omni/deploy/bagel.yaml
-        commands:
-          - |
-            timeout 40m bash -c "
-              export VLLM_IMAGE_FETCH_TIMEOUT=60
-              pytest -s -v tests/e2e/online_serving/test_bagel.py -m 'core_model' --run-level 'core_model'
-            "
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "Diffusion · Wan22 Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_wan22_t2v.py
-          - vllm_omni/diffusion/models/wan2_2/
-        commands:
-          - |
-            timeout 40m bash -c "
-              export VLLM_IMAGE_FETCH_TIMEOUT=60
-              pytest -s -v tests/e2e/online_serving/test_wan22_t2v.py -m 'core_model' --run-level 'core_model'
-            "
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "Diffusion · Cosmos3 Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_cosmos3.py
-          - vllm_omni/diffusion/models/cosmos3/
-        commands:
-          - |
-            timeout 40m bash -c "
-              export VLLM_IMAGE_FETCH_TIMEOUT=60
-              pytest -s -v tests/e2e/online_serving/test_cosmos3.py -m 'core_model' --run-level 'core_model'
-            "
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-
-      - label: "Diffusion · Qwen Image Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen_image.py
-          - vllm_omni/diffusion/models/qwen_image/
-        commands:
-          - |
-            timeout 40m bash -c "
-              export VLLM_IMAGE_FETCH_TIMEOUT=60
-              pytest -s -v tests/e2e/online_serving/test_qwen_image.py -m 'core_model' --run-level 'core_model'
-            "
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "Diffusion · Qwen Image Layered Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen_image_layered.py
-          - vllm_omni/diffusion/models/qwen_image/
-        commands:
-          - |
-            timeout 40m bash -c "
-              export VLLM_IMAGE_FETCH_TIMEOUT=60
-              pytest -s -v tests/e2e/online_serving/test_qwen_image_layered.py -m 'core_model' --run-level 'core_model'
-            "
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "Diffusion · GGUF Plugin Test"
-        soft_fail: true
-        source_file_dependencies:
-          - tests/plugin/test_gguf_diffusion.py
-          - vllm_omni/diffusion/model_loader
-        commands:
-          - pip install "vllm-gguf-plugin>=0.0.3"
-          - timeout 30m pytest -sv tests/plugin/test_gguf_diffusion.py
-        agents:
-          queue: "gpu_1_queue" # g6.4xlarge instance on AWS, has 1 L4 GPU
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"

.buildkite/test-weekly.yml

@@ -1,451 +0,0 @@
-env:
-  VLLM_WORKER_MULTIPROC_METHOD: spawn
-  VLLM_USE_FLASHINFER_MOE_FP16: 0
-  HF_HUB_DOWNLOAD_TIMEOUT: 300
-  HF_HUB_ETAG_TIMEOUT: 60
-
-steps:
-  - label: "Reliability Test - qwen3-omni"
-    timeout_in_minutes: 180
-    depends_on: upload-weekly-pipeline
-    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
-    commands:
-      - pytest -s -v tests/dfx/reliability/test_reliability_qwen3_omni.py -m "slow"
-    agents:
-      queue: "mithril-h100-pool"
-    plugins:
-      - kubernetes:
-          podSpec:
-            containers:
-              - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                resources:
-                  limits:
-                    nvidia.com/gpu: 2
-                volumeMounts:
-                  - name: devshm
-                    mountPath: /dev/shm
-                  - name: hf-cache
-                    mountPath: /root/.cache/huggingface
-                env:
-                  - name: HF_HOME
-                    value: /root/.cache/huggingface
-                  - name: HF_TOKEN
-                    valueFrom:
-                      secretKeyRef:
-                        name: hf-token-secret
-                        key: token
-            nodeSelector:
-              node.kubernetes.io/instance-type: gpu-h100-sxm
-            volumes:
-              - name: devshm
-                emptyDir:
-                  medium: Memory
-              - name: hf-cache
-                hostPath:
-                  path: /mnt/hf-cache
-                  type: DirectoryOrCreate
-
-  - label: "Reliability Test - wan22"
-    timeout_in_minutes: 180
-    depends_on: upload-weekly-pipeline
-    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
-    commands:
-      - pytest -s -v tests/dfx/reliability/test_reliability_wan22.py -m "slow"
-    agents:
-      queue: "mithril-h100-pool"
-    plugins:
-      - kubernetes:
-          podSpec:
-            containers:
-              - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                resources:
-                  limits:
-                    nvidia.com/gpu: 2
-                volumeMounts:
-                  - name: devshm
-                    mountPath: /dev/shm
-                  - name: hf-cache
-                    mountPath: /root/.cache/huggingface
-                env:
-                  - name: HF_HOME
-                    value: /root/.cache/huggingface
-                  - name: HF_TOKEN
-                    valueFrom:
-                      secretKeyRef:
-                        name: hf-token-secret
-                        key: token
-            nodeSelector:
-              node.kubernetes.io/instance-type: gpu-h100-sxm
-            volumes:
-              - name: devshm
-                emptyDir:
-                  medium: Memory
-              - name: hf-cache
-                hostPath:
-                  path: /mnt/hf-cache
-                  type: DirectoryOrCreate
-
-  - label: "Reliability Test - hunyuan-image"
-    timeout_in_minutes: 180
-    depends_on: upload-weekly-pipeline
-    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
-    commands:
-      - pytest -s -v tests/dfx/reliability/test_reliability_hunyuan_image.py -m "slow"
-    agents:
-      queue: "mithril-h100-pool"
-    plugins:
-      - kubernetes:
-          podSpec:
-            containers:
-              - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                resources:
-                  limits:
-                    nvidia.com/gpu: 4
-                volumeMounts:
-                  - name: devshm
-                    mountPath: /dev/shm
-                  - name: hf-cache
-                    mountPath: /root/.cache/huggingface
-                env:
-                  - name: HF_HOME
-                    value: /root/.cache/huggingface
-                  - name: HF_TOKEN
-                    valueFrom:
-                      secretKeyRef:
-                        name: hf-token-secret
-                        key: token
-            nodeSelector:
-              node.kubernetes.io/instance-type: gpu-h100-sxm
-            volumes:
-              - name: devshm
-                emptyDir:
-                  medium: Memory
-              - name: hf-cache
-                hostPath:
-                  path: /mnt/hf-cache
-                  type: DirectoryOrCreate
-
-  - label: "Reliability Test - voxcpm2"
-    timeout_in_minutes: 180
-    depends_on: upload-weekly-pipeline
-    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
-    commands:
-      - pytest -s -v tests/dfx/reliability/test_reliability_voxcpm2.py -m "slow"
-    agents:
-      queue: "gpu_1_queue"
-    plugins:
-      - docker#v5.2.0:
-          image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-          always-pull: true
-          propagate-environment: true
-          shm-size: "8gb"
-          environment:
-            - "HF_HOME=/fsx/hf_cache"
-            - "HF_TOKEN"
-          volumes:
-            - "/fsx/hf_cache:/fsx/hf_cache"
-
-  # Invalid HTTP / WS payloads vs live Omni servers (Issue #3649 tracks known skips).
-  - group: "Reliability Test - Invalid parameters Test"
-    depends_on: upload-weekly-pipeline
-    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
-    steps:
-      - label: "Invalid parameters Test · H100"
-        timeout_in_minutes: 120
-        commands:
-          - pytest -s -v tests/dfx/reliability/invalid_param_test/ -m "slow and H100"
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 2
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      # Stable Audio Open, Qwen3-TTS speech/batch/voices paths marked ``L4``.
-      - label: "Invalid parameters Test · L4"
-        timeout_in_minutes: 120
-        commands:
-          - pytest -s -v tests/dfx/reliability/invalid_param_test/ -m "slow and L4"
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-  - group: ":card_index_dividers: E2E Tests"
-    depends_on: upload-weekly-pipeline
-    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
-    steps:
-      - label: "Omni · L4"
-        timeout_in_minutes: 90
-        commands:
-          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
-          - pytest -s -v tests/e2e/*/test_qwen2_5_omni_expansion.py --run-level "core_model"
-          - >-
-            pytest -s -v tests/e2e/
-            --ignore=tests/e2e/online_serving/test_qwen2_5_omni_expansion.py
-            --ignore=tests/e2e/offline_inference/test_qwen2_5_omni_expansion.py
-            -m "slow and L4 and omni"
-            --run-level "full_model"
-        agents:
-          queue: "gpu_4_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Omni · H100"
-        timeout_in_minutes: 90
-        commands:
-          - >-
-            pytest -sv
-            tests/e2e/ -m "slow and H100 and omni"
-            --run-level "full_model"
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 2
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "TTS · L4"
-        timeout_in_minutes: 90
-        commands:
-          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
-          - export VLLM_USE_DEEP_GEMM="0"
-          - export VLLM_MOE_USE_DEEP_GEMM="0"
-          - |
-            failed=0
-            for f in $(pytest --collect-only -q tests/e2e/ -m "slow and L4 and tts" --run-level "full_model" | grep '\.py::' | sed -E 's|::.*||' | sort -u); do
-              pytest -s -v "$$f" -m "slow and L4 and tts" --run-level "full_model" || failed=1
-            done
-            exit $$failed
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "TTS · H100"
-        timeout_in_minutes: 90
-        commands:
-          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
-          - export VLLM_USE_DEEP_GEMM="0"
-          - export VLLM_MOE_USE_DEEP_GEMM="0"
-          - >-
-            pytest -sv
-            tests/e2e/ -m "slow and H100 and tts"
-            --run-level "full_model"
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "Diffusion · L4"
-        timeout_in_minutes: 60
-        commands:
-          - >-
-            pytest -sv
-            tests/e2e/ -m "slow and diffusion and L4"
-            --run-level "full_model"
-        agents:
-          queue: "gpu_4_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-                - "HF_TOKEN"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
-
-      - label: "Diffusion · H100 · Single-GPU"
-        timeout_in_minutes: 120
-        commands:
-          - >-
-            pytest -sv
-            tests/e2e/ -m "slow and diffusion and H100 and not distributed_cuda"
-            --run-level "full_model"
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 1
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate
-
-      - label: "Diffusion · H100 · Multi-GPU"
-        timeout_in_minutes: 90
-        commands:
-          - >-
-            pytest -sv
-            tests/e2e/ -m "slow and diffusion and H100 and distributed_cuda"
-            --run-level "full_model"
-        agents:
-          queue: "mithril-h100-pool"
-        plugins:
-          - kubernetes:
-              podSpec:
-                containers:
-                  - image: 936637512419.dkr.ecr.us-west-2.amazonaws.com/vllm-ci-pull-through-cache/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-                    resources:
-                      limits:
-                        nvidia.com/gpu: 2
-                    volumeMounts:
-                      - name: devshm
-                        mountPath: /dev/shm
-                      - name: hf-cache
-                        mountPath: /root/.cache/huggingface
-                    env:
-                      - name: HF_HOME
-                        value: /root/.cache/huggingface
-                      - name: HF_TOKEN
-                        valueFrom:
-                          secretKeyRef:
-                            name: hf-token-secret
-                            key: token
-                nodeSelector:
-                  node.kubernetes.io/instance-type: gpu-h100-sxm
-                volumes:
-                  - name: devshm
-                    emptyDir:
-                      medium: Memory
-                  - name: hf-cache
-                    hostPath:
-                      path: /mnt/hf-cache
-                      type: DirectoryOrCreate

.claude/skills/add-diffusion-model/SKILL.md

@@ -317,7 +317,7 @@ Required updates:
 
 ### Step 8: Add E2E Tests
 
-**Follow the [vllm-omni-test skill](../vllm-omni-test/SKILL.md)** for markers, file naming, Buildkite wiring, and run commands. Also read [l4_functionality_tests.inc.md](https://github.com/vllm-project/vllm-omni/blob/main/docs/contributing/ci/test_examples/l4_functionality_tests.inc.md) and [CI_5levels.md](https://github.com/vllm-project/vllm-omni/blob/main/docs/contributing/ci/CI_5levels.md).
+**Follow the [vllm-omni-test skill](../vllm-omni-test/SKILL.md)** for markers, file naming, Buildkite wiring, and run commands. Also read [l4_functionality_tests.inc.md](https://github.com/vllm-project/vllm-omni/blob/main/docs/contributing/ci/test_examples/l4_functionality_tests.inc.md), [test_system_overview.md](https://github.com/vllm-project/vllm-omni/blob/main/docs/contributing/ci/test_system_overview.md), and [test_writing_guide.md](https://github.com/vllm-project/vllm-omni/blob/main/docs/contributing/ci/test_writing_guide.md).
 
 Classify the model's **CI priority** first:
 
@@ -418,7 +418,8 @@ def enable_cache_for_your_model(pipeline, cache_config):
 
 #### 9d. Register the custom enabler
 
-Add your enabler to `CUSTOM_DIT_ENABLERS` in `vllm_omni/diffusion/cache/cache_dit_backend.py`:
+Add your enabler to `CUSTOM_DIT_ENABLERS` in
+`vllm_omni/diffusion/cache/cachedit/model_specific.py`:
 
 ```python
 CUSTOM_DIT_ENABLERS = {

.claude/skills/add-diffusion-model/references/cache-dit-patterns.md

@@ -29,8 +29,13 @@ vLLM-Omni integrates cache-dit through `CacheDiTBackend`:
 | `cache_dit.refresh_context()` | Updates cache context when `num_inference_steps` changes |
 
 **Source files**:
-- `vllm_omni/diffusion/cache/cache_dit_backend.py` — `CacheDiTBackend`, enablers, `CUSTOM_DIT_ENABLERS`
-- `vllm_omni/diffusion/cache/` — cache backend implementations
+- `vllm_omni/diffusion/cache/cachedit/__init__.py` — public package API
+- `vllm_omni/diffusion/cache/cachedit/backend.py` — generic backend lifecycle and integration
+- `vllm_omni/diffusion/cache/cachedit/config.py` — Cache-DiT configuration types
+- `vllm_omni/diffusion/cache/cachedit/model_specific.py` — model-specific adapters, enablers, and registry
+
+Outside `vllm_omni/diffusion/cache`, import Cache-DiT symbols only from
+`vllm_omni.diffusion.cache.cachedit`.
 
 ## Standard Models: Automatic Support
 
@@ -175,7 +180,8 @@ Inspect your block's `forward()` return type and residual connection pattern to
 
 ## Registering Custom Enablers
 
-Add your enabler to `CUSTOM_DIT_ENABLERS` in `vllm_omni/diffusion/cache/cache_dit_backend.py`:
+Add your enabler to `CUSTOM_DIT_ENABLERS` in
+`vllm_omni/diffusion/cache/cachedit/model_specific.py`:
 
 ```python
 CUSTOM_DIT_ENABLERS = {
@@ -248,7 +254,7 @@ Models listed in `_NO_CACHE_ACCELERATION` in `vllm_omni/diffusion/registry.py` d
 
 | Model | Path | Notes |
 |-------|------|-------|
-| Standard DiT | `cache_dit_backend.py::enable_cache_for_dit` | Default enabler, automatic |
-| Wan2.2 | `cache_dit_backend.py::enable_cache_for_wan22` | Dual-transformer, auto-detects mode |
-| LongCat | `cache_dit_backend.py::enable_cache_for_longcat_image` | Multi-block-list |
-| BAGEL | `cache_dit_backend.py::enable_cache_for_bagel` | Complex omni model |
+| Standard DiT | `cachedit.backend::enable_cache_for_dit` | Default enabler, automatic |
+| Wan2.2 | `cachedit.model_specific::enable_cache_for_wan22` | Dual-transformer, auto-detects mode |
+| LongCat | `cachedit.config::CacheDiTAdapterConfig` | Declarative multi-block-list adapter |
+| BAGEL | `cachedit.model_specific::BagelCachedAdapter` | Custom cached adapter |

.claude/skills/add-tts-model/SKILL.md

@@ -307,7 +307,7 @@ def build_voice_clone_prompt(ref_audio_path: str, text: str, codec) -> list:
 
 ### Test Case Writing (CI Levels)
 
-**Follow the [vllm-omni-test skill](../vllm-omni-test/SKILL.md)** for markers, file naming (`test_{slug}.py` / `test_{slug}_expansion.py`), Buildkite wiring, and copy-paste run commands. Also read [CI_5levels.md](https://github.com/vllm-project/vllm-omni/blob/main/docs/contributing/ci/CI_5levels.md) and [tests_style.md](https://github.com/vllm-project/vllm-omni/blob/main/docs/contributing/ci/tests_style.md).
+**Follow the [vllm-omni-test skill](../vllm-omni-test/SKILL.md)** for markers, file naming (`test_{slug}.py` / `test_{slug}_expansion.py`), Buildkite wiring, and copy-paste run commands. Also read [test_system_overview.md](https://github.com/vllm-project/vllm-omni/blob/main/docs/contributing/ci/test_system_overview.md) and [test_writing_guide.md](https://github.com/vllm-project/vllm-omni/blob/main/docs/contributing/ci/test_writing_guide.md).
 
 Classify the model's **CI priority** first (high / medium / low). High-priority TTS models are typically those on the integration hot path or listed in tracking issues such as [#1832](https://github.com/vllm-project/vllm-omni/issues/1832); medium and low tiers cover the long tail. When unsure, ask the reviewer which tier applies.
 

.claude/skills/vllm-omni-test/SKILL.md

@@ -9,7 +9,7 @@ description: Generate and run tests for vllm-project/vllm-omni with CI-aligned l
 
 Use this skill to generate minimal, stable test cases and run them with the correct marker/level strategy for [vllm-project/vllm-omni](https://github.com/vllm-project/vllm-omni).
 
-**Link convention:** Paths such as `.buildkite/` and `docs/contributing/` live at the **vllm-omni repo root**. Markdown links use repo-relative paths from this skill file (e.g. `../../../.buildkite/test-ready.yml`, `../../../docs/contributing/ci/CI_5levels.md`).
+**Link convention:** Paths such as `.buildkite/` and `docs/contributing/` live at the **vllm-omni repo root**. Markdown links use repo-relative paths from this skill file (e.g. `../../../.buildkite/cuda/test-ready.yml`, `../../../docs/contributing/ci/test_system_overview.md`).
 
 Default priorities:
 
@@ -96,7 +96,7 @@ Existing references: `tests/e2e/offline_inference/test_qwen2_5_omni.py` (L2-styl
 |----------|------------------|---------------------------|---------------------------|
 | **Offline inference e2e** | `tests/e2e/offline_inference/` | **Module (default):** `omni_runner` + `omni_runner_handler`. **Function (isolation only):** `omni_runner_function` + `omni_runner_handler_function`. Diffusion/TTS may use `Omni(...).generate` directly | L2: `core_model` + **one of** `omni` / `tts` / `diffusion`; `@hardware_test(...)` when GPU/NPU is required |
 | **Online serving e2e** | `tests/e2e/online_serving/` | **Module (default):** `omni_server` + `openai_client`. **Function (isolation only):** `omni_server_function` + `openai_client_function`. Clients: `send_omni_request` (omni), `send_audio_speech_request` (tts), `send_diffusion_request` / `send_video_diffusion_request` / `send_images_generations_request` (diffusion) | Baseline smoke: **`core_model` + `advanced_model`**; heavier paths: `advanced_model` only; L4 expansion: `full_model` |
-| **Documentation / runnable examples** | `tests/examples/offline_inference/`, `tests/examples/online_serving/` | **Offline docs (preferred):** extract Python/Bash blocks from the doc README (e.g. `ReadmeSnippet.extract_readme_snippets`), `pytest.mark.parametrize` each snippet, run via `example_runner.run` with a stable `output_subfolder`. **Online docs:** copy client/request scripts into dedicated tests and keep them in sync with the doc page. | Usually **L4**: `advanced_model`, often `example` plus hardware marks matching the nightly docs-example job (see `.buildkite/test-nightly.yml`). Full conventions: [docs/contributing/ci/test_examples/l4_doc_example_tests.inc.md](../../../docs/contributing/ci/test_examples/l4_doc_example_tests.inc.md) (introduced in [PR #1910](https://github.com/vllm-project/vllm-omni/pull/1910): naming, output directory layout, skip rules, avoid trimming `num_inference_steps` without a strong CI reason). |
+| **Documentation / runnable examples** | `tests/examples/offline_inference/`, `tests/examples/online_serving/` | **Offline docs (preferred):** extract Python/Bash blocks from the doc README (e.g. `ReadmeSnippet.extract_readme_snippets`), `pytest.mark.parametrize` each snippet, run via `example_runner.run` with a stable `output_subfolder`. **Online docs:** copy client/request scripts into dedicated tests and keep them in sync with the doc page. | Usually **L4**: `advanced_model`, often `example` plus hardware marks matching the nightly docs-example job (see `.buildkite/cuda/test-nightly.yml`). Full conventions: [docs/contributing/ci/test_examples/l4_doc_example_tests.inc.md](../../../docs/contributing/ci/test_examples/l4_doc_example_tests.inc.md) (introduced in [PR #1910](https://github.com/vllm-project/vllm-omni/pull/1910): naming, output directory layout, skip rules, avoid trimming `num_inference_steps` without a strong CI reason). |
 | **Performance / benchmark** | `tests/dfx/perf/tests/*.json` + `run_*_benchmark.py` | JSON or script-driven server + load config; assert explicit metrics / baselines | L4 Perf: JSON `mark` with `full_model` + `omni`/`tts`/`diffusion`; wire `test-nightly.yml` Perf steps |
 | **Invalid parameter / negative HTTP validation** | `tests/dfx/reliability/invalid_param_test/` | Live `omni_server` + low-level `send_*_http_request` with `err_code` / `err_message` | `pytest.mark.slow` + `omni` / `tts` / `diffusion` + `@hardware_marks` (`H100` or `L4`); CI in **`test-weekly.yml`** (not ready/merge/nightly) |
 
@@ -460,7 +460,7 @@ When the user asks for **invalid parameter validation**, **invalid request bodie
 
 4. **Adding a new model** to invalid-param coverage: append a `pytest.param(OmniServerParams(model="...", stage_config_path=..., server_args=...), ...)` entry to the script’s `_PARAMS` list — do **not** create `test_invalid_<model>.py` unless the route is new.
 
-5. **CI — `.buildkite/test-weekly.yml` only** (not `test-ready.yml` / `test-merge.yml` / `test-nightly.yml`):
+5. **CI — `.buildkite/cuda/test-weekly.yml` only** (not `test-ready.yml` / `test-merge.yml` / `test-nightly.yml`):
 
 | Weekly step | Command | When your cases run |
 |-------------|---------|---------------------|
@@ -559,7 +559,7 @@ L1 tests under `tests/entrypoints/` may keep **minimal** asserts next to the han
 
 #### Omni Test Writing Guidance (L1-L4 Layering)
 
-When the goal is a general Omni/multimodal test case, prioritize mapping the test to the correct purpose, directory, and resource assumptions aligned with the layers (see [CI_5levels.md](../../../docs/contributing/ci/CI_5levels.md)):
+When the goal is a general Omni/multimodal test case, prioritize mapping the test to the correct purpose, directory, and resource assumptions aligned with the layers (see [test_system_overview.md](../../../docs/contributing/ci/test_system_overview.md) and [test_writing_guide.md](../../../docs/contributing/ci/test_writing_guide.md)):
 
 - **L1**: Unit/logic validation on CPU (`core_model and cpu`). Cover input validation, branches, and exception paths (`tests/<component>/test_*.py`). **Mock with `mocker` / `monkeypatch` only — not `unittest.mock`.**
 - **L2**: Basic e2e (online/offline basic scenarios). Prefer dummy/lightweight models to validate the end-to-end request-to-output-structure/streaming chain (typically `tests/e2e/online_serving/` and `tests/e2e/offline_inference/`).
@@ -629,10 +629,10 @@ If the test is not already collected by an existing pipeline command (for exampl
 
 | Test level | Edit this file | Typical trigger / intent |
 |------------|----------------|---------------------------|
-| **L1** and **L2** | [`.buildkite/test-ready.yml`](../../../.buildkite/test-ready.yml) | PR **ready** label; L1 CPU + L2 GPU/basic e2e (steps labeled **Omni ·**, **TTS ·**, **Diffusion ·**) |
-| **L3** | [`.buildkite/test-merge.yml`](../../../.buildkite/test-merge.yml) | Post-merge; `advanced_model` integration per model type |
-| **L4** | [`.buildkite/test-nightly.yml`](../../../.buildkite/test-nightly.yml) | Nightly; grouped by model type (see below) |
-| **Invalid param / reliability (weekly)** | [`.buildkite/test-weekly.yml`](../../../.buildkite/test-weekly.yml) | Weekly; `tests/dfx/reliability/invalid_param_test/` — **not** L1–L4 e2e pipelines |
+| **L1** and **L2** | [`.buildkite/cuda/test-ready.yml`](../../../.buildkite/cuda/test-ready.yml) | PR **ready** label; L1 CPU + L2 GPU/basic e2e (steps labeled **Omni ·**, **TTS ·**, **Diffusion ·**) |
+| **L3** | [`.buildkite/cuda/test-merge.yml`](../../../.buildkite/cuda/test-merge.yml) | Post-merge; `advanced_model` integration per model type |
+| **L4** | [`.buildkite/cuda/test-nightly.yml`](../../../.buildkite/cuda/test-nightly.yml) | Nightly; grouped by model type (see below) |
+| **Invalid param / reliability (weekly)** | [`.buildkite/cuda/test-weekly.yml`](../../../.buildkite/cuda/test-weekly.yml) | Weekly; `tests/dfx/reliability/invalid_param_test/` — **not** L1–L4 e2e pipelines |
 
 **Level-specific Buildkite delivery (follow the user’s requested level):**
 
@@ -809,21 +809,21 @@ Use **repo-relative paths** (no leading `./`). Prefer **directory** entries (`..
 
 When extending an **existing** E2E step to run an additional test file, **append** that file (and any new model/deploy paths) to `source_file_dependencies` on the same step. When authoring a PR, mention the dependency list in the **Buildkite change** section of your test plan.
 
-#### L3 / L4: validating Buildkite from a feature branch (root `pipeline.yml`)
+#### L3 / L4: validating Buildkite from a feature branch (`cuda/pipeline.yml`)
 
-Production behavior is driven by [`.buildkite/pipeline.yml`](../../../.buildkite/pipeline.yml): it builds the CI image, then uploads **L2** (`test-ready.yml` on non-`main`), **L3** (`test-merge.yml` when `build.branch == "main"`), and **L4** (`test-nightly.yml` when `build.env("NIGHTLY") == "1"`).
+Production behavior is driven by [`.buildkite/cuda/pipeline.yml`](../../../.buildkite/cuda/pipeline.yml): it builds the CI image, then uploads **L2** (`test-ready.yml` on non-`main`), **L3** (`test-merge.yml` when `build.branch == "main"`), and **L4** (`test-nightly.yml` when `build.env("NIGHTLY") == "1"`).
 
 When you add or debug **L3** or **L4** tests and need those child pipelines to run **off `main`** or **without** `NIGHTLY=1`, apply **temporary** edits (revert before merge):
 
 1. **Point the upload at the right file (optional shortcut)**  
-   In the **Upload Ready Pipeline** step, the command is `buildkite-agent pipeline upload .buildkite/test-ready.yml`. For a one-off run you may change that path to `.buildkite/test-merge.yml` (L3) or `.buildkite/test-nightly.yml` (L4), and adjust the step’s `if:` so it does not conflict with the other upload steps (avoid double-uploading unless intentional).
+   In the **Upload Ready Pipeline** step, the command is `buildkite-agent pipeline upload .buildkite/cuda/test-ready.yml`. For a one-off run you may change that path to `.buildkite/cuda/test-merge.yml` (L3) or `.buildkite/cuda/test-nightly.yml` (L4), and adjust the step’s `if:` so it does not conflict with the other upload steps (avoid double-uploading unless intentional).
 
 2. **L3 — merge pipeline**  
-   The gate `if: build.branch == "main"` lives on the **Upload Merge Pipeline** step in `pipeline.yml` (not inside `test-merge.yml`). **Comment out that `if` line** so `test-merge.yml` is uploaded on your feature branch. Alternatively rely on step (1) instead of the dedicated merge upload step.
+   The gate `if: build.branch == "main"` lives on the **Upload Merge Pipeline** step in `cuda/pipeline.yml` (not inside `test-merge.yml`). **Comment out that `if` line** so `test-merge.yml` is uploaded on your feature branch. Alternatively rely on step (1) instead of the dedicated merge upload step.
 
 3. **L4 — nightly pipeline**  
-   **Comment out** `if: build.env("NIGHTLY") == "1"` on the **Upload Nightly Pipeline** step in `pipeline.yml` so the nightly definition is uploaded without setting the env var.  
-   Child steps in [`test-nightly.yml`](../../../.buildkite/test-nightly.yml) often repeat `if: build.env("NIGHTLY") == "1"`; **comment out those lines on the steps you need to run**, otherwise they will still be skipped after upload.
+   **Comment out** `if: build.env("NIGHTLY") == "1"` on the **Upload Nightly Pipeline** step in `cuda/pipeline.yml` so the nightly definition is uploaded without setting the env var.  
+   Child steps in [`test-nightly.yml`](../../../.buildkite/cuda/test-nightly.yml) often repeat `if: build.env("NIGHTLY") == "1"`; **comment out those lines on the steps you need to run**, otherwise they will still be skipped after upload.
 
 ### Step 6: Run Tests
 
@@ -890,5 +890,5 @@ When completing a request, return:
 ## Additional Resources
 
 - Marker and command routing: [references/test-routing.md](references/test-routing.md)
-- CI pipelines (vllm-omni): [test-ready.yml](../../../.buildkite/test-ready.yml) (L1/L2), [test-merge.yml](../../../.buildkite/test-merge.yml) (L3), [test-nightly.yml](../../../.buildkite/test-nightly.yml) (L4), [test-weekly.yml](../../../.buildkite/test-weekly.yml) (invalid param / reliability)
+- CI pipelines (vllm-omni): [test-ready.yml](../../../.buildkite/cuda/test-ready.yml) (L1/L2), [test-merge.yml](../../../.buildkite/cuda/test-merge.yml) (L3), [test-nightly.yml](../../../.buildkite/cuda/test-nightly.yml) (L4), [test-weekly.yml](../../../.buildkite/cuda/test-weekly.yml) (invalid param / reliability)
 - L4 documentation example tests (naming, extraction vs copied scripts, output dirs, skips): [docs/contributing/ci/test_examples/l4_doc_example_tests.inc.md](../../../docs/contributing/ci/test_examples/l4_doc_example_tests.inc.md) — see also [PR #1910](https://github.com/vllm-project/vllm-omni/pull/1910)

.claude/skills/vllm-omni-test/references/test-routing.md

@@ -2,7 +2,7 @@
 
 Use this reference to map testing goals to levels, markers, and runnable commands.
 
-**Repo paths** (`.buildkite/`, `docs/contributing/…`): link with repo-relative paths from this file (e.g. `../../../../.buildkite/test-ready.yml`, `../../../../docs/contributing/ci/CI_5levels.md`).
+**Repo paths** (`.buildkite/`, `docs/contributing/…`): link with repo-relative paths from this file (e.g. `../../../../.buildkite/cuda/test-ready.yml`, `../../../../docs/contributing/ci/test_system_overview.md`).
 
 ## Model-centric e2e filename convention (L2–L4)
 
@@ -224,7 +224,7 @@ When the user’s test plan includes **invalid parameter validation**, **invalid
 | Sleep / wakeup / server control | `test_invalid_server_control.py` |
 
 3. **Style:** `pytestmark = [pytest.mark.slow, pytest.mark.<type>]`; `_PARAMS` + `hardware_marks`; `send_*_http_request` with `err_code` + `err_message`; parametrized `body_spec` rows with `id=`; `_minimal_*_json()` helpers; `_SKIP_ISSUE_3649` when tracked in [#3649](https://github.com/vllm-project/vllm-omni/issues/3649).
-4. **CI:** [`.buildkite/test-weekly.yml`](../../../../.buildkite/test-weekly.yml) group **Reliability Test - Invalid parameters Test** — **not** ready/merge/nightly.
+4. **CI:** [`.buildkite/cuda/test-weekly.yml`](../../../../.buildkite/cuda/test-weekly.yml) group **Reliability Test - Invalid parameters Test** — **not** ready/merge/nightly.
 
 ```bash
 # Weekly H100 (diffusion / omni / video invalid-param)
@@ -291,10 +291,10 @@ When adding or modifying tests, do not stop at “where the file lives” — al
 
 | Level | Repo file | Model-type grouping |
 |-------|-----------|---------------------|
-| L1, L2 | [`.buildkite/test-ready.yml`](../../../../.buildkite/test-ready.yml) | Steps prefixed **Omni ·**, **TTS ·**, **Diffusion ·** under **E2E Test** — **`source_file_dependencies` required** |
-| L3 | [`.buildkite/test-merge.yml`](../../../../.buildkite/test-merge.yml) | Per-model E2E steps; `-m "advanced_model and …"` — **`source_file_dependencies` required** |
-| L4 | [`.buildkite/test-nightly.yml`](../../../../.buildkite/test-nightly.yml) | **Omni / TTS / Diffusion X2I(&A&T) / X2V** — each group may have **Function**, **Accuracy**, **Perf**, **Doc** steps |
-| Invalid param (weekly) | [`.buildkite/test-weekly.yml`](../../../../.buildkite/test-weekly.yml) | **Invalid parameters Test · H100** / **· L4** — sweeps `tests/dfx/reliability/invalid_param_test/` |
+| L1, L2 | [`.buildkite/cuda/test-ready.yml`](../../../../.buildkite/cuda/test-ready.yml) | Steps prefixed **Omni ·**, **TTS ·**, **Diffusion ·** under **E2E Test** — **`source_file_dependencies` required** |
+| L3 | [`.buildkite/cuda/test-merge.yml`](../../../../.buildkite/cuda/test-merge.yml) | Per-model E2E steps; `-m "advanced_model and …"` — **`source_file_dependencies` required** |
+| L4 | [`.buildkite/cuda/test-nightly.yml`](../../../../.buildkite/cuda/test-nightly.yml) | **Omni / TTS / Diffusion X2I(&A&T) / X2V** — each group may have **Function**, **Accuracy**, **Perf**, **Doc** steps |
+| Invalid param (weekly) | [`.buildkite/cuda/test-weekly.yml`](../../../../.buildkite/cuda/test-weekly.yml) | **Invalid parameters Test · H100** / **· L4** — sweeps `tests/dfx/reliability/invalid_param_test/` |
 
 ### `source_file_dependencies` (E2E Test only — ready & merge)
 
@@ -380,7 +380,7 @@ Common patterns in `test-ready.yml` / `test-merge.yml`:
 
 When extending an existing step (e.g. add `tests/e2e/offline_inference/test_qwen_image.py` to **Diffusion · Qwen Image Test**), update `source_file_dependencies` and `commands` only; **keep the existing `agents` + `plugins` block unchanged** unless the hardware tier changes.
 
-The root [`.buildkite/pipeline.yml`](../../../../.buildkite/pipeline.yml) decides **which** child file is uploaded. To run **L3** on a feature branch, comment out `if: build.branch == "main"` on the merge upload step. To run **L4** without `NIGHTLY=1`, comment out the nightly upload step’s `if: build.env("NIGHTLY") == "1"` and the same `if` on relevant steps inside `test-nightly.yml`. Revert such edits before merging.
+[`.buildkite/cuda/pipeline.yml`](../../../../.buildkite/cuda/pipeline.yml) decides **which** child file is uploaded. To run **L3** on a feature branch, comment out `if: build.branch == "main"` on the merge upload step. To run **L4** without `NIGHTLY=1`, comment out the nightly upload step’s `if: build.env("NIGHTLY") == "1"` and the same `if` on relevant steps inside `test-nightly.yml`. Revert such edits before merging.
 
 Platform-specific pipelines (e.g. AMD) follow the same level → file pairing under `.buildkite/`.
 

.github/workflows/pre-commit.yml

@@ -24,3 +24,9 @@ jobs:
           python-version: "3.12"
           cache: 'pip'
       - uses: pre-commit/action@v3.0.1
+        # Don't run the pytest mark check against everything in the CI
+        # for now since a lot of files are still missing them. This check
+        # is used to help contributors remember to add the marks when they
+        # commit.
+        env:
+          SKIP: check-test-ci-coverage

.gitignore

@@ -259,7 +259,6 @@ configs/production.yaml
 configs/development.yaml
 
 # Docker
-.dockerignore
 Dockerfile.dev
 discussion
 tmp_test

.pre-commit-config.yaml

@@ -54,6 +54,13 @@ repos:
       verbose: true
       stages: [commit-msg]
 
+    - id: check-test-ci-coverage
+      name: Ensure test files have pytestmarks
+      entry: python tools/pre_commit/check_test_marks.py
+      language: python
+      types: [python]
+      files: ^tests/
+
     # Keep `suggestion` last
     - id: suggestion
       name: Suggestion

benchmarks/README.md

@@ -28,6 +28,14 @@ Benchmarks for GLM-Image performance across HuggingFace baseline, vLLM-Omni offl
 - **Tasks**: text-to-image and image-to-image
 - **Key metrics**: request/image throughput, latency percentiles, optional per-stage pipeline timings
 
+### [LingBot-Video](lingbot_video/README.md) — Dense and MoE Parity
+
+Manual cross-runtime validation for the dense LingBot-Video pipeline and the
+LingBot-Video MoE transformer.
+
+- **Dense pipeline**: decoded-video MAE, MSE, PSNR, latency, and optional steady-state timings
+- **MoE transformer**: bitwise router, sparse-block, shared-expert, and full-transformer parity
+
 ### [Distributed](distributed/omni_connectors/README.md) — RDMA Connector Testing
 
 RDMA environment setup and transfer tests for `MooncakeTransferEngineConnector`, including pytest-based single-node checks and manual cross-node benchmarks.

benchmarks/lingbot_video/README.md

@@ -0,0 +1,62 @@
+# LingBot-Video Benchmarks
+
+These manual benchmarks compare the upstream LingBot-Video implementation with
+vLLM-Omni. Run all commands from the repository root.
+
+| Benchmark | Script | Purpose |
+|-----------|--------|---------|
+| Dense pipeline parity | `dense_pipeline_parity.py` | Compare generated videos and report MAE, MSE, PSNR, latency, and optional steady-state timings |
+| MoE numerical parity | `moe_transformer_parity.py` | Check bitwise parity for the sparse MoE block and the full MoE transformer |
+
+Both benchmarks require a local LingBot-Video checkout and local model files.
+They do not run in CI.
+
+## Dense pipeline parity
+
+The dense benchmark launches the upstream Diffusers inference script and the
+vLLM-Omni offline example with identical prompts and sampling parameters. It
+then compares the decoded MP4 frames.
+
+```bash
+CUDA_VISIBLE_DEVICES=0 \
+python benchmarks/lingbot_video/dense_pipeline_parity.py \
+  --model /path/to/lingbot-video-dense-1.3b \
+  --official-repo /path/to/lingbot-video \
+  --output-dir /tmp/lingbot_dense_parity
+```
+
+Use `--runs N` to additionally measure repeated in-process vLLM-Omni
+requests. The video comparison reports MAE, MSE, and PSNR; it is an end-to-end
+pipeline diagnostic rather than a bitwise transformer check.
+
+## MoE numerical parity
+
+The lightweight block comparison uses deterministic weights and inputs to
+cover correction-bias routing, group-limited top-k, routed experts, FP32
+scatter-weighted restore, padding masks, and the shared expert.
+
+```bash
+CUDA_VISIBLE_DEVICES=0 \
+python benchmarks/lingbot_video/moe_transformer_parity.py \
+  --scope block \
+  --official-repo /path/to/lingbot-video \
+  --output-json /tmp/lingbot_moe_block_parity.json
+```
+
+The full comparison loads the upstream and vLLM-Omni 30B transformers
+sequentially and compares their output tensors. It requires a CUDA device with
+`torch._grouped_mm` and enough memory for one 30B BF16 transformer.
+
+```bash
+CUDA_VISIBLE_DEVICES=0 \
+python benchmarks/lingbot_video/moe_transformer_parity.py \
+  --scope transformer \
+  --official-repo /path/to/lingbot-video \
+  --model /path/to/lingbot-video-moe-30b-a3b \
+  --output-json /tmp/lingbot_moe_transformer_parity.json
+```
+
+For the correctness oracle, the upstream transformer uses
+`diffusers:_native_math` and vLLM-Omni uses
+`TORCH_SDPA + SDPBackend.MATH`. The command exits with a nonzero status unless
+the selected comparisons are bitwise equal.

benchmarks/lingbot_video/dense_pipeline_parity.py

@@ -0,0 +1,217 @@
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
+
+from __future__ import annotations
+
+import argparse
+import json
+import re
+import statistics
+import subprocess
+import sys
+import time
+from pathlib import Path
+from typing import Any
+
+import imageio.v3 as iio
+import numpy as np
+import torch
+
+from vllm_omni.diffusion.data import DiffusionParallelConfig
+from vllm_omni.entrypoints.omni import Omni
+from vllm_omni.inputs.data import OmniDiffusionSamplingParams
+from vllm_omni.platforms import current_omni_platform
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description="LingBot dense parity and timing harness.")
+    parser.add_argument("--model", default="/home/models/lingbot-video-dense-1.3b")
+    parser.add_argument("--official-repo", default="/tmp/lingbot-video")
+    parser.add_argument("--output-dir", default="/tmp/lingbot_dense_parity")
+    parser.add_argument("--prompt", default="a robotic arm picks up a red block")
+    parser.add_argument("--height", type=int, default=192)
+    parser.add_argument("--width", type=int, default=320)
+    parser.add_argument("--num-frames", type=int, default=9)
+    parser.add_argument("--steps", type=int, default=2)
+    parser.add_argument("--guidance-scale", type=float, default=3.0)
+    parser.add_argument("--shift", type=float, default=3.0)
+    parser.add_argument("--seed", type=int, default=42)
+    parser.add_argument("--fps", type=int, default=24)
+    parser.add_argument("--runs", type=int, default=0, help="Optional in-process steady-state native runs.")
+    return parser.parse_args()
+
+
+def _run(cmd: list[str], cwd: Path | None = None) -> tuple[float, str]:
+    start = time.perf_counter()
+    proc = subprocess.run(cmd, cwd=cwd, text=True, stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
+    elapsed = time.perf_counter() - start
+    if proc.returncode != 0:
+        print(proc.stdout)
+        raise subprocess.CalledProcessError(proc.returncode, proc.args, output=proc.stdout)
+    return elapsed, proc.stdout
+
+
+def _read_video(path: Path) -> np.ndarray:
+    frames = iio.imread(path, plugin="pyav")
+    if frames.dtype != np.float32:
+        frames = frames.astype(np.float32) / 255.0
+    return frames
+
+
+def _compare(a: np.ndarray, b: np.ndarray) -> dict[str, Any]:
+    if a.shape != b.shape:
+        return {"shape_a": list(a.shape), "shape_b": list(b.shape), "shape_match": False}
+    diff = a - b
+    mse = float(np.mean(diff * diff))
+    mae = float(np.mean(np.abs(diff)))
+    psnr = float("inf") if mse == 0 else float(20.0 * np.log10(1.0 / np.sqrt(mse)))
+    return {
+        "shape": list(a.shape),
+        "shape_match": True,
+        "mae": mae,
+        "mse": mse,
+        "psnr": psnr,
+    }
+
+
+def _extract_total_generation_seconds(log: str) -> float | None:
+    match = re.search(r"Total generation time:\s*([0-9.]+)\s*seconds", log)
+    return float(match.group(1)) if match else None
+
+
+def _peak_memory_mb(output: Any) -> float:
+    result = output[0] if isinstance(output, list) and output else output
+    return float(getattr(result, "peak_memory_mb", 0.0) or 0.0)
+
+
+def _benchmark_native_requests(args: argparse.Namespace) -> dict[str, Any]:
+    load_start = time.perf_counter()
+    omni = Omni(
+        model=args.model,
+        model_class_name="LingBotVideoPipeline",
+        flow_shift=args.shift,
+        parallel_config=DiffusionParallelConfig(),
+    )
+    load_seconds = time.perf_counter() - load_start
+    prompt = {"prompt": args.prompt}
+    latencies: list[float] = []
+    peak_mb = 0.0
+
+    for _ in range(args.runs):
+        generator = torch.Generator(device=current_omni_platform.device_type).manual_seed(args.seed)
+        sampling_params = OmniDiffusionSamplingParams(
+            height=args.height,
+            width=args.width,
+            num_frames=args.num_frames,
+            num_inference_steps=args.steps,
+            guidance_scale=args.guidance_scale,
+            generator=generator,
+            output_type="pt",
+            extra_args={"shift": args.shift},
+        )
+        start = time.perf_counter()
+        output = omni.generate(prompt, sampling_params)
+        latencies.append(time.perf_counter() - start)
+        peak_mb = max(peak_mb, _peak_memory_mb(output))
+        del output
+    omni.close()
+
+    return {
+        "load_seconds": load_seconds,
+        "run_seconds": latencies,
+        "mean_seconds": statistics.mean(latencies) if latencies else None,
+        "median_seconds": statistics.median(latencies) if latencies else None,
+        "min_seconds": min(latencies) if latencies else None,
+        "max_seconds": max(latencies) if latencies else None,
+        "peak_mb": peak_mb,
+    }
+
+
+def main() -> None:
+    args = parse_args()
+    output_dir = Path(args.output_dir)
+    output_dir.mkdir(parents=True, exist_ok=True)
+    official_out = output_dir / "official_diffusers.mp4"
+    native_out = output_dir / "vllm_omni_native.mp4"
+
+    official_cmd = [
+        sys.executable,
+        str(Path(args.official_repo) / "scripts" / "inference.py"),
+        "--backend",
+        "diffusers",
+        "--model_dir",
+        args.model,
+        "--mode",
+        "t2v",
+        "--prompt",
+        args.prompt,
+        "--output",
+        str(official_out),
+        "--height",
+        str(args.height),
+        "--width",
+        str(args.width),
+        "--num_frames",
+        str(args.num_frames),
+        "--steps",
+        str(args.steps),
+        "--guidance_scale",
+        str(args.guidance_scale),
+        "--shift",
+        str(args.shift),
+        "--seed",
+        str(args.seed),
+        "--fps",
+        str(args.fps),
+        "--quiet_progress",
+    ]
+    native_cmd = [
+        sys.executable,
+        "examples/offline_inference/text_to_video/text_to_video_lingbot.py",
+        "--model",
+        args.model,
+        "--prompt",
+        args.prompt,
+        "--output",
+        str(native_out),
+        "--height",
+        str(args.height),
+        "--width",
+        str(args.width),
+        "--num-frames",
+        str(args.num_frames),
+        "--num-inference-steps",
+        str(args.steps),
+        "--guidance-scale",
+        str(args.guidance_scale),
+        "--flow-shift",
+        str(args.shift),
+        "--seed",
+        str(args.seed),
+        "--fps",
+        str(args.fps),
+    ]
+
+    official_time, official_log = _run(official_cmd, cwd=Path(args.official_repo))
+    native_time, native_log = _run(native_cmd, cwd=Path.cwd())
+
+    result = {
+        "official_seconds": official_time,
+        "native_process_seconds": native_time,
+        "native_request_seconds": _extract_total_generation_seconds(native_log),
+        "official_output": str(official_out),
+        "native_output": str(native_out),
+        "accuracy": _compare(_read_video(official_out), _read_video(native_out)),
+        "official_log_tail": official_log.splitlines()[-20:],
+        "native_log_tail": native_log.splitlines()[-20:],
+    }
+    if args.runs > 0:
+        result["native_steady_state"] = _benchmark_native_requests(args)
+
+    result_path = output_dir / "comparison.json"
+    result_path.write_text(json.dumps(result, indent=2, sort_keys=True) + "\n")
+    print(json.dumps(result, indent=2, sort_keys=True))
+
+
+if __name__ == "__main__":
+    main()

benchmarks/lingbot_video/moe_transformer_parity.py

@@ -0,0 +1,399 @@
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
+
+"""Reproduce LingBot-Video MoE block and transformer numerical parity."""
+
+from __future__ import annotations
+
+import argparse
+import gc
+import importlib
+import json
+import os
+import sys
+import time
+from pathlib import Path
+from typing import Any
+
+import torch
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument(
+        "--scope",
+        choices=("block", "transformer", "all"),
+        default="block",
+        help="Run the lightweight sparse-block check, the real transformer check, or both.",
+    )
+    parser.add_argument(
+        "--official-repo",
+        type=Path,
+        required=True,
+        help="Local checkout of https://github.com/Robbyant/lingbot-video.",
+    )
+    parser.add_argument(
+        "--model",
+        default=None,
+        help="Local MoE checkpoint path or a checkpoint already present in the Hugging Face cache.",
+    )
+    parser.add_argument("--transformer-subfolder", default="transformer")
+    parser.add_argument("--output-json", type=Path, default=None)
+    parser.add_argument("--device", default="cuda")
+    parser.add_argument("--seed", type=int, default=42)
+    parser.add_argument("--num-frames", type=int, default=1)
+    parser.add_argument("--height", type=int, default=8)
+    parser.add_argument("--width", type=int, default=8)
+    parser.add_argument("--text-length", type=int, default=16)
+    return parser.parse_args()
+
+
+def _load_official_module(official_repo: Path):
+    transformer_file = official_repo / "lingbot_video" / "transformer_lingbot_video.py"
+    if not transformer_file.is_file():
+        raise FileNotFoundError(f"Upstream transformer not found: {transformer_file}")
+    sys.path.insert(0, str(official_repo))
+    module = importlib.import_module("lingbot_video.transformer_lingbot_video")
+    module_path = Path(module.__file__).resolve()
+    if official_repo.resolve() not in module_path.parents:
+        raise RuntimeError(f"Imported LingBot module from {module_path}, not {official_repo.resolve()}")
+    return module
+
+
+def _load_native_module():
+    return importlib.import_module("vllm_omni.diffusion.models.lingbot_video.lingbot_video_transformer")
+
+
+def _tensor_metrics(actual: torch.Tensor, expected: torch.Tensor) -> dict[str, Any]:
+    actual_float = actual.float()
+    expected_float = expected.float()
+    diff = (actual_float - expected_float).abs()
+    denominator = actual_float.norm() * expected_float.norm()
+    cosine = (
+        float(torch.dot(actual_float.flatten(), expected_float.flatten()) / denominator)
+        if float(denominator) > 0
+        else 1.0
+    )
+    return {
+        "shape": list(actual.shape),
+        "equal": bool(torch.equal(actual, expected)),
+        "max_abs": float(diff.max()) if diff.numel() else 0.0,
+        "mean_abs": float(diff.mean()) if diff.numel() else 0.0,
+        "rmse": float(torch.sqrt(torch.mean(diff.square()))) if diff.numel() else 0.0,
+        "cosine": cosine,
+        "finite": bool(torch.isfinite(actual).all() and torch.isfinite(expected).all()),
+    }
+
+
+def _initialize_block(module: torch.nn.Module, seed: int) -> None:
+    generator = torch.Generator(device="cpu").manual_seed(seed)
+    with torch.no_grad():
+        for parameter in module.parameters():
+            values = torch.randn(parameter.shape, generator=generator, dtype=torch.float32)
+            parameter.copy_((values * 0.02).to(parameter.dtype))
+        module.router.e_score_correction_bias.copy_(torch.tensor([0.9, 0.8, 1.0, 0.0, 0.7, 0.6, 0.5, 0.4]))
+
+
+def _apply_padding(
+    scores: torch.Tensor,
+    padding_mask: torch.Tensor,
+    route_scale: float,
+) -> torch.Tensor:
+    scores = scores * padding_mask.unsqueeze(-1).to(scores.dtype)
+    scores = scores / (scores.sum(dim=-1, keepdim=True) + 1e-9)
+    return scores * route_scale
+
+
+def _run_block_parity(
+    official_module,
+    native_module,
+    *,
+    device: torch.device,
+    seed: int,
+) -> dict[str, Any]:
+    if not hasattr(torch, "_grouped_mm"):
+        raise RuntimeError("The block parity check requires torch._grouped_mm.")
+
+    common = {
+        "hidden_size": 16,
+        "num_experts": 8,
+        "top_k": 2,
+        "moe_intermediate_size": 8,
+        "score_func": "sigmoid",
+        "norm_topk_prob": True,
+        "n_group": 4,
+        "topk_group": 1,
+        "routed_scaling_factor": 1.5,
+        "n_shared_experts": 1,
+    }
+    official = official_module.LingBotVideoSparseMoeBlock(
+        intermediate_size=32,
+        **common,
+    )
+    native = native_module.LingBotVideoSparseMoeBlock(**common)
+    _initialize_block(official, seed)
+    native.load_state_dict(official.state_dict(), strict=True)
+    official = official.to(device=device, dtype=torch.bfloat16).eval()
+    native = native.to(device=device, dtype=torch.bfloat16).eval()
+
+    generator = torch.Generator(device=device).manual_seed(seed)
+    hidden_states = torch.randn(
+        2,
+        5,
+        common["hidden_size"],
+        generator=generator,
+        device=device,
+        dtype=torch.bfloat16,
+    )
+    padding_mask = torch.tensor(
+        [1, 1, 1, 1, 1, 1, 1, 1, 0, 0],
+        device=device,
+        dtype=torch.float32,
+    )
+    tokens = hidden_states.reshape(-1, common["hidden_size"])
+
+    with torch.inference_mode():
+        official_router = official.router(tokens)
+        official_indices, official_scores = official_router[:2]
+        native_indices, native_scores = native.router(tokens)
+
+        official_scores = _apply_padding(
+            official_scores,
+            padding_mask,
+            official.router.route_scale,
+        )
+        native_scores = _apply_padding(
+            native_scores,
+            padding_mask,
+            native.router.route_scale,
+        )
+        official_routed = official._run_selected_experts(
+            tokens,
+            official_scores,
+            official_indices,
+        )
+        native_routed = native._run_selected_experts(
+            tokens,
+            native_scores,
+            native_indices,
+        )
+        official_shared = official.shared_experts(hidden_states)
+        native_shared = native.shared_experts(hidden_states)
+        official_output = official(hidden_states, padding_mask=padding_mask)
+        native_output = native(hidden_states, padding_mask=padding_mask)
+
+    result = {
+        "router_indices_equal": bool(torch.equal(native_indices, official_indices)),
+        "router_scores": _tensor_metrics(native_scores, official_scores),
+        "routed_output": _tensor_metrics(native_routed, official_routed),
+        "shared_output": _tensor_metrics(native_shared, official_shared),
+        "final_output": _tensor_metrics(native_output, official_output),
+    }
+    result["exact"] = bool(
+        result["router_indices_equal"]
+        and result["router_scores"]["equal"]
+        and result["routed_output"]["equal"]
+        and result["shared_output"]["equal"]
+        and result["final_output"]["equal"]
+    )
+    return result
+
+
+def _release_cuda() -> None:
+    gc.collect()
+    torch.accelerator.empty_cache()
+    torch.accelerator.synchronize()
+
+
+def _make_transformer_inputs(
+    config,
+    args: argparse.Namespace,
+) -> dict[str, torch.Tensor]:
+    generator = torch.Generator(device="cpu").manual_seed(args.seed)
+    hidden_states = torch.randn(
+        1,
+        int(config.in_channels),
+        args.num_frames,
+        args.height,
+        args.width,
+        generator=generator,
+        dtype=torch.float32,
+    ).to(torch.bfloat16)
+    encoder_hidden_states = torch.randn(
+        1,
+        args.text_length,
+        int(config.text_dim),
+        generator=generator,
+        dtype=torch.float32,
+    ).to(torch.bfloat16)
+    return {
+        "hidden_states": hidden_states,
+        "timestep": torch.tensor([500.0], dtype=torch.float32),
+        "encoder_hidden_states": encoder_hidden_states,
+        "encoder_attention_mask": torch.ones(1, args.text_length, dtype=torch.long),
+    }
+
+
+def _load_transformer(
+    transformer_cls,
+    args: argparse.Namespace,
+    device: torch.device,
+):
+    start = time.perf_counter()
+    model = transformer_cls.from_pretrained(
+        args.model,
+        subfolder=args.transformer_subfolder,
+        torch_dtype=torch.bfloat16,
+        local_files_only=True,
+        low_cpu_mem_usage=True,
+    )
+    model = model.to(device=device, dtype=torch.bfloat16).eval()
+    torch.accelerator.synchronize()
+    return model, time.perf_counter() - start
+
+
+def _forward_transformer(
+    model,
+    inputs: dict[str, torch.Tensor],
+    device: torch.device,
+    attention_context,
+) -> tuple[torch.Tensor, float, float]:
+    torch.accelerator.reset_peak_memory_stats(device)
+    gpu_inputs = {name: value.to(device) for name, value in inputs.items()}
+    start = time.perf_counter()
+    with torch.inference_mode(), attention_context:
+        output = model(**gpu_inputs, return_dict=False)[0]
+    torch.accelerator.synchronize()
+    elapsed = time.perf_counter() - start
+    peak_gib = torch.accelerator.max_memory_reserved(device) / (1024**3)
+    return output.cpu(), elapsed, peak_gib
+
+
+def _run_transformer_parity(
+    official_module,
+    native_module,
+    *,
+    args: argparse.Namespace,
+    device: torch.device,
+) -> dict[str, Any]:
+    if args.model is None:
+        raise ValueError("--model is required for transformer parity.")
+
+    from diffusers.models.attention_dispatch import attention_backend
+    from torch.nn.attention import SDPBackend, sdpa_kernel
+
+    from vllm_omni.diffusion.config import set_current_diffusion_config
+    from vllm_omni.diffusion.data import AttentionConfig, AttentionSpec, OmniDiffusionConfig
+
+    official, official_load = _load_transformer(
+        official_module.LingBotVideoTransformer3DModel,
+        args,
+        device,
+    )
+    inputs = _make_transformer_inputs(official.config, args)
+    official_output, official_forward, official_peak = _forward_transformer(
+        official,
+        inputs,
+        device,
+        attention_backend("_native_math"),
+    )
+    del official
+    _release_cuda()
+
+    native_config = OmniDiffusionConfig(
+        diffusion_attention_config=AttentionConfig(
+            default=AttentionSpec(backend="TORCH_SDPA"),
+        ),
+    )
+    with set_current_diffusion_config(native_config):
+        native, native_load = _load_transformer(
+            native_module.LingBotVideoTransformer3DModel,
+            args,
+            device,
+        )
+    native_backend_prefs = {block.attn.attn.backend_pref for block in native.blocks}
+    if native_backend_prefs != {"TORCH_SDPA"}:
+        raise RuntimeError(
+            f"Failed to force the native transformer to TORCH_SDPA: resolved preferences were {native_backend_prefs}."
+        )
+    native_output, native_forward, native_peak = _forward_transformer(
+        native,
+        inputs,
+        device,
+        sdpa_kernel(SDPBackend.MATH),
+    )
+    del native
+    _release_cuda()
+
+    metrics = _tensor_metrics(native_output, official_output)
+    return {
+        "backends": {
+            "official": "diffusers:_native_math",
+            "native": "TORCH_SDPA+SDPBackend.MATH",
+        },
+        "official": {
+            "load_seconds": official_load,
+            "forward_seconds": official_forward,
+            "peak_reserved_gib": official_peak,
+        },
+        "native": {
+            "load_seconds": native_load,
+            "forward_seconds": native_forward,
+            "peak_reserved_gib": native_peak,
+        },
+        "output": metrics,
+        "exact": metrics["equal"],
+    }
+
+
+def main() -> int:
+    args = parse_args()
+    device = torch.device(args.device)
+    if device.type != "cuda" or not torch.cuda.is_available():
+        raise RuntimeError("LingBot MoE parity requires an available CUDA device.")
+
+    os.environ["DIFFUSERS_ATTN_BACKEND"] = "_native_math"
+    os.environ["DIFFUSION_ATTENTION_BACKEND"] = "TORCH_SDPA"
+    os.environ["LINGBOT_MOE_EXPERT_BACKEND"] = "grouped_mm"
+    os.environ["LINGBOT_MOE_PAD_BACKEND"] = "loop"
+    os.environ["LINGBOT_MOE_REORDER_BACKEND"] = "sort"
+    os.environ["LINGBOT_MOE_RESTORE_BACKEND"] = "scatter"
+
+    official_module = _load_official_module(args.official_repo)
+    native_module = _load_native_module()
+
+    result: dict[str, Any] = {
+        "settings": {
+            "scope": args.scope,
+            "seed": args.seed,
+            "device": str(device),
+            "official_repo": str(args.official_repo.resolve()),
+            "model": args.model,
+        }
+    }
+    if args.scope in {"block", "all"}:
+        result["block"] = _run_block_parity(
+            official_module,
+            native_module,
+            device=device,
+            seed=args.seed,
+        )
+    if args.scope in {"transformer", "all"}:
+        result["transformer"] = _run_transformer_parity(
+            official_module,
+            native_module,
+            args=args,
+            device=device,
+        )
+
+    result["exact"] = all(section["exact"] for name, section in result.items() if name in {"block", "transformer"})
+    payload = json.dumps(result, indent=2, sort_keys=True)
+    print(payload)
+    if args.output_json is not None:
+        args.output_json.parent.mkdir(parents=True, exist_ok=True)
+        args.output_json.write_text(payload + "\n", encoding="utf-8")
+    return 0 if result["exact"] else 1
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

docker/Dockerfile.ci

@@ -1,12 +1,26 @@
-# Target vllm version: v0.24.0 (releases/v0.24.0 branch).
-# The pinned VLLM_PRECOMPILED_WHEEL_COMMIT and VLLM_BASE_TAG below must
-# be updated on every rebase to match the new upstream vllm commit/tag.
+# Target vllm version: v0.26.0 (releases/v0.26.0 branch).
+# Keep VLLM_BASE_TAG aligned with the target vLLM release on every rebase.
 ARG VLLM_BASE_IMAGE=vllm/vllm-openai
-ARG VLLM_BASE_TAG=v0.25.0
-FROM ${VLLM_BASE_IMAGE}:${VLLM_BASE_TAG}
-ARG APP_DIR=/workspace/vllm-omni
+ARG VLLM_BASE_TAG=v0.26.0
+
+# We use registry-based layer caching for the dependencies,
+# where the tag is determined based on the hash of the contents
+# of files like pyproject.toml, etc. The deps stage should have
+# everything except the actual project dir, since that should not
+# change much across branches.
+FROM ${VLLM_BASE_IMAGE}:${VLLM_BASE_TAG} AS deps
+# Keep APP_DIR outside /workspace so k8s Buildkite workspace mounts cannot
+# shadow the editable install target (#5364). Docker WORKDIR creates /opt if
+# needed.
+ARG APP_DIR=/opt/vllm-omni
 WORKDIR ${APP_DIR}
-COPY . .
+
+RUN ln -sf /usr/bin/python3 /usr/bin/python
+
+# Placeholder package tree so editable install can register APP_DIR.
+# The final stage COPY --link fills this directory with the real sources;
+# the editable .pth / egg-link keeps resolving imports there — no PYTHONPATH.
+RUN mkdir -p vllm_omni && touch vllm_omni/__init__.py
 
 # Install system dependencies
 RUN apt-get update && \
@@ -14,9 +28,30 @@ RUN apt-get update && \
     apt-get clean && \
     rm -rf /var/lib/apt/lists/*
 
-RUN uv pip install --system ".[dev]"
+# Copy dependency related things first and install deps/
+# We set the version str to 0.0.0 since it doesn't matter
+# for CI
+COPY pyproject.toml setup.py ./
+COPY requirements/ requirements/
+ENV SETUPTOOLS_SCM_PRETEND_VERSION=0.0.0
+# Editable install: deps + project metadata are cached in this layer, while
+# the package itself is linked to APP_DIR (populated later by COPY --link).
+# Do NOT use a non-editable install here — that copies an empty stub into
+# site-packages and breaks ``import vllm_omni.entrypoints``.
+RUN uv pip install --system -e ".[dev]"
 
+# The final stage just copies the working dir on top of the
+# dependencies image. This uses --link so that we can skip
+# the image pull in Buildkite when we have a cache hit.
+#
+# IMPORTANT: Do not make any modifications that depend on the
+# file system after the --link copy, otherwise we'd need to pull
+# the heavy base image to build it. Editable install above already
+# registered APP_DIR, so we must not re-pip-install after COPY.
+FROM deps
+ARG APP_DIR=/opt/vllm-omni
+WORKDIR ${APP_DIR}
 
-RUN ln -sf /usr/bin/python3 /usr/bin/python
+COPY --link . .
 
 ENTRYPOINT []

docker/Dockerfile.ci.dockerignore

@@ -0,0 +1,37 @@
+# NOTE: this is largely taken from vLLM
+
+# Exclude .git for CI builds — version is set via SETUPTOOLS_SCM_PRETEND_VERSION
+.git
+.venv
+/build
+dist
+
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+.mypy_cache
+
+# Distribution / packaging
+.Python
+/build/
+cmake-build-*/
+CMakeUserPresets.json
+develop-eggs/
+/dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+rust/target/

docker/Dockerfile.cuda

@@ -1,4 +1,4 @@
-ARG BASE_IMAGE=vllm/vllm-openai:v0.25.0
+ARG BASE_IMAGE=vllm/vllm-openai:v0.26.0
 FROM ${BASE_IMAGE}
 
 ARG COMMON_WORKDIR=/app

docker/Dockerfile.npu.ci

@@ -37,6 +37,13 @@ RUN source /usr/local/Ascend/ascend-toolkit/set_env.sh && \
       "pyttsx3>=2.99" \
       "opencv-python>=4.12.0.88" \
       "imageio[ffmpeg]>=0.6.0" \
+      "jiwer>=3.0.0" \
+      "zhon>=2.0.0" \
+      "zhconv>=1.4.2" \
+      "funasr>=1.0.0" \
+      "hyperpyyaml>=1.2.0" \
+      "onnx>=1.16.0" \
+      "librosa>=0.10.0" \
       --index-url http://cache-service.nginx-pypi-cache.svc.cluster.local/pypi/simple \
       --trusted-host cache-service.nginx-pypi-cache.svc.cluster.local \
       --default-timeout=60 && \
@@ -45,6 +52,17 @@ RUN source /usr/local/Ascend/ascend-toolkit/set_env.sh && \
       --trusted-host cache-service.nginx-pypi-cache.svc.cluster.local \
       --default-timeout=60
 
+# MiniCPM-o 4.5 Stage-2 vocoder assets: s3tokenizer plus flashcosyvoice, which
+# is shipped by step-audio2. --no-deps keeps pip from replacing the Ascend torch
+# build and from shadowing onnxruntime-cann with the vanilla onnxruntime wheel;
+# the deps these two actually need are installed above.
+RUN python3 -m pip install -v --no-deps \
+      "s3tokenizer>=0.3.0" \
+      "step-audio2>=1.0.0" \
+      --index-url http://cache-service.nginx-pypi-cache.svc.cluster.local/pypi/simple \
+      --trusted-host cache-service.nginx-pypi-cache.svc.cluster.local \
+      --default-timeout=60
+
 ENV VLLM_WORKER_MULTIPROC_METHOD=spawn
 
 ENTRYPOINT []

docker/Dockerfile.npu.ci.a3

@@ -37,6 +37,13 @@ RUN source /usr/local/Ascend/ascend-toolkit/set_env.sh && \
       "pyttsx3>=2.99" \
       "opencv-python>=4.12.0.88" \
       "imageio[ffmpeg]>=0.6.0" \
+      "jiwer>=3.0.0" \
+      "zhon>=2.0.0" \
+      "zhconv>=1.4.2" \
+      "funasr>=1.0.0" \
+      "hyperpyyaml>=1.2.0" \
+      "onnx>=1.16.0" \
+      "librosa>=0.10.0" \
       --index-url http://cache-service.nginx-pypi-cache.svc.cluster.local/pypi/simple \
       --trusted-host cache-service.nginx-pypi-cache.svc.cluster.local \
       --default-timeout=60 && \
@@ -45,6 +52,17 @@ RUN source /usr/local/Ascend/ascend-toolkit/set_env.sh && \
       --trusted-host cache-service.nginx-pypi-cache.svc.cluster.local \
       --default-timeout=60
 
+# MiniCPM-o 4.5 Stage-2 vocoder assets: s3tokenizer plus flashcosyvoice, which
+# is shipped by step-audio2. --no-deps keeps pip from replacing the Ascend torch
+# build and from shadowing onnxruntime-cann with the vanilla onnxruntime wheel;
+# the deps these two actually need are installed above.
+RUN python3 -m pip install -v --no-deps \
+      "s3tokenizer>=0.3.0" \
+      "step-audio2>=1.0.0" \
+      --index-url http://cache-service.nginx-pypi-cache.svc.cluster.local/pypi/simple \
+      --trusted-host cache-service.nginx-pypi-cache.svc.cluster.local \
+      --default-timeout=60
+
 ENV VLLM_WORKER_MULTIPROC_METHOD=spawn
 
 ENTRYPOINT []

docker/Dockerfile.rocm

@@ -1,4 +1,4 @@
-ARG BASE_IMAGE=vllm/vllm-openai-rocm:v0.25.0
+ARG BASE_IMAGE=vllm/vllm-openai-rocm:v0.26.0
 FROM ${BASE_IMAGE} AS base
 
 # Declare a variable to know if we want to use the nightly build or the stable build.
@@ -10,14 +10,14 @@ FROM ${BASE_IMAGE} AS base
 #    we should swap over to use stable release ASAP.
 #    We should avoid relying on custom commits.
 ARG USE_NIGHTLY_BUILD=0
-ARG VLLM_VERSION_OR_COMMIT_HASH=89138b21cc246ae944c741d5c399c148e2b770ab
+ARG VLLM_VERSION_OR_COMMIT_HASH=v0.26.0
 ARG ARG_PYTORCH_ROCM_ARCH
 ENV PYTORCH_ROCM_ARCH=${ARG_PYTORCH_ROCM_ARCH:-${PYTORCH_ROCM_ARCH}}
 
 ARG COMMON_WORKDIR=/app
 WORKDIR ${COMMON_WORKDIR}
 
-# Step 1: Setup - Install system dependencies
+# Step 1: Setup - Install system dependencies.
 # Need to include ffmpeg because vllm rocm upstream docker image
 # does not include it.
 RUN apt-get update && \
@@ -38,6 +38,26 @@ RUN if [ "${USE_NIGHTLY_BUILD}" = "1" ]; then \
         rm -rf vllm/.git; \
     fi
 
+# Build TorchCodec after any optional nightly vLLM reinstall so it targets the
+# final torch/ROCm ABI. Keep this layer before COPY . so source changes do not
+# invalidate the expensive native build. FFmpeg development headers are
+# removed after the wheel is built.
+COPY tools/install_torchcodec_rocm.sh /tmp/install_torchcodec_rocm.sh
+RUN --mount=type=cache,target=/root/.cache/pip \
+    --mount=type=cache,target=/root/.cache/torchcodec-wheels \
+    apt-get update && \
+    apt-get install -y --no-install-recommends \
+        pkg-config \
+        libavcodec-dev libavformat-dev libavutil-dev \
+        libswscale-dev libavdevice-dev libavfilter-dev libswresample-dev && \
+    bash /tmp/install_torchcodec_rocm.sh && \
+    apt-get purge -y \
+        libavcodec-dev libavformat-dev libavutil-dev \
+        libswscale-dev libavdevice-dev libavfilter-dev libswresample-dev && \
+    rm -f /tmp/install_torchcodec_rocm.sh && \
+    apt-get clean && \
+    rm -rf /var/lib/apt/lists/*
+
 # Step 3: Copy vllm-omni code and install without uv
 RUN mkdir -p ${COMMON_WORKDIR}/vllm-omni
 COPY . ${COMMON_WORKDIR}/vllm-omni
@@ -59,7 +79,10 @@ RUN printf '%s\n' \
     '    os._exit(_exit_code)' \
     > ${COMMON_WORKDIR}/vllm-omni/conftest.py
 
-RUN cd ${COMMON_WORKDIR}/vllm-omni && uv pip install --python "$(python3 -c 'import sys; print(sys.executable)')" --no-cache-dir ".[dev]" --no-build-isolation
+RUN cd ${COMMON_WORKDIR}/vllm-omni && \
+    uv pip install --python "$(python3 -c 'import sys; print(sys.executable)')" \
+        --no-cache-dir ".[dev]" --no-build-isolation && \
+    python3 -c "from torchcodec.decoders import VideoDecoder"
 
 RUN ln -sf /usr/bin/python3 /usr/bin/python
 

docker/Dockerfile.xpu

@@ -88,7 +88,7 @@ ENV UV_EXTRA_INDEX_URL=${PIP_EXTRA_INDEX_URL}
 ENV UV_INDEX_STRATEGY="unsafe-best-match"
 ENV UV_LINK_MODE="copy"
 
-ARG VLLM_VERSION=v0.25.0
+ARG VLLM_VERSION=v0.26.0
 RUN git clone -b ${VLLM_VERSION} https://github.com/vllm-project/vllm
 WORKDIR /workspace/vllm
 

docs/.nav.yml

@@ -16,13 +16,11 @@ nav:
   - Examples:
     - examples/README.md
     - Offline Inference:
-      - BAGEL-7B-MoT: user_guide/examples/offline_inference/bagel.md
       - GLM-Image Multistage End-to-End Inference: user_guide/examples/offline_inference/glm_image.md
       - HunyuanImage-3.0 Image-to-Text Inference: user_guide/examples/offline_inference/hunyuan_image3.md
       - Image-To-Image: user_guide/examples/offline_inference/image_to_image.md
       - Image-To-Video: user_guide/examples/offline_inference/image_to_video.md
       - InternVLA-A1: user_guide/examples/offline_inference/internvla_a1.md
-      - MammothModa2-Preview: user_guide/examples/offline_inference/mammothmodal2_preview.md
       - MiMo-Audio: user_guide/examples/offline_inference/mimo_audio.md
       - Qwen2.5-Omni: user_guide/examples/offline_inference/qwen2_5_omni.md
       - Qwen3-Omni: user_guide/examples/offline_inference/qwen3_omni.md
@@ -56,6 +54,7 @@ nav:
     - Diffusion Features:
       - Overview: user_guide/diffusion_features.md
       - Feature Compatibility: user_guide/feature_compatibility.md
+      - Regional Compilation: user_guide/diffusion/regional_compilation.md
       - Cache Acceleration:
         - TeaCache: user_guide/diffusion/cache_acceleration/teacache.md
         - Cache-DiT: user_guide/diffusion/cache_acceleration/cache_dit.md
@@ -117,6 +116,7 @@ nav:
       - design/feature/hsdp.md
       - design/feature/cache_dit.md
       - design/feature/teacache.md
+      - design/feature/skip_softmax.md
       - design/feature/async_chunk.md
       - design/feature/vae_parallel.md
       - design/feature/diffusion_step_execution.md

docs/api/README.md

@@ -62,7 +62,6 @@ Engine classes for offline and online inference.
 
 Core scheduling and caching components.
 
-- [vllm_omni.core.sched.omni_ar_scheduler.KVCacheTransferData][]
 - [vllm_omni.core.sched.omni_ar_scheduler.OmniARScheduler][]
 - [vllm_omni.core.sched.omni_generation_scheduler.OmniGenerationScheduler][]
 - [vllm_omni.core.sched.output.OmniCachedRequestData][]
@@ -131,7 +130,7 @@ Worker classes and model runners for distributed inference.
 - [vllm_omni.diffusion.worker.diffusion_worker.DiffusionWorker][]
 - [vllm_omni.diffusion.worker.diffusion_worker.WorkerProc][]
 - [vllm_omni.diffusion.worker.diffusion_worker.WorkerWrapperBase][]
-- [vllm_omni.diffusion.worker.utils.DiffusionRequestState][]
+- [vllm_omni.diffusion.worker.utils.StepRequestState][]
 - [vllm_omni.diffusion.worker.utils.RunnerOutput][]
 - [vllm_omni.platforms.npu.worker.npu_ar_model_runner.ExecuteModelState][]
 - [vllm_omni.platforms.npu.worker.npu_ar_model_runner.NPUARModelRunner][]

docs/cli/serve.md

@@ -8,8 +8,7 @@ The stage-based CLI is designed for deployments that require launching each pipe
 - For **migrated models** that utilize the bundled deployment YAML configurations located in
   `vllm_omni/deploy/`, the `--deploy-config` flag is only required to override the default configuration. By default, executing `vllm serve MODEL --omni ...`
   automatically loads the bundled deployment configuration.
-- For **legacy models** utilizing configuration files located in
-  `vllm_omni/model_executor/stage_configs/`, the `--stage-configs-path` parameter remains mandatory.
+- For custom legacy `stage_args` YAMLs, pass the file with `--stage-configs-path`.
 
 Example: Initializing Stage 0 (Orchestrator and API Server):
 The commands below show a common device mapping where Stage 0 uses GPU 0 and

docs/configuration/stage_configs.md

@@ -3,7 +3,7 @@
 In vLLM-Omni, a model's `PipelineConfig` defines its fixed stage topology, while a deploy configuration controls how those stages run. Models that have not migrated to this split still use legacy stage configurations with a `stage_args` schema.
 
 !!! note
-    Default deploy config YAMLs (for example, `vllm_omni/deploy/qwen2_5_omni.yaml`, `vllm_omni/deploy/qwen3_omni_moe.yaml`, and `vllm_omni/deploy/qwen3_tts.yaml`) are bundled and loaded automatically when neither `--stage-configs-path` nor `--deploy-config` is provided — the model registry resolves the right pipeline + deploy YAML by `model_type`. The bundled defaults have been verified on 1xH100 for Qwen2.5-Omni and 2xH100 for Qwen3-Omni. Models that have not yet migrated to the new schema continue to use the legacy `vllm_omni/model_executor/stage_configs/<model>.yaml` files via `--stage-configs-path`.
+    Default deploy config YAMLs (for example, `vllm_omni/deploy/qwen2_5_omni.yaml`, `vllm_omni/deploy/qwen3_omni_moe.yaml`, and `vllm_omni/deploy/qwen3_tts.yaml`) are bundled and loaded automatically when neither `--stage-configs-path` nor `--deploy-config` is provided — the model registry resolves the right pipeline + deploy YAML by `model_type`. Custom legacy `stage_args` YAMLs are still accepted via `--stage-configs-path`.
 
 ## Deploy configuration schema
 
@@ -65,7 +65,7 @@ connectors:
 
 | Connector class | Use case | `extra` keys |
 |-----------------|----------|--------------|
-| `SharedMemoryConnector` | Same-host KV transfer between stages (default for bundled YAMLs). | `shm_threshold_bytes` (int, default `65536`). |
+| `SharedMemoryConnector` | Same-host KV transfer between stages (default for bundled YAMLs). | None. All payloads use shared memory. |
 | `MooncakeStoreConnector` | Cross-host KV transfer over TCP. Required for multi-node deployments. | `host`, `metadata_server`, `master`, `segment` (int bytes), `localbuf` (int bytes), `proto` (`"tcp"` / `"rdma"`). |
 
 A stage references a connector by name in its `input_connectors` / `output_connectors`:
@@ -377,7 +377,12 @@ Default: `"0"`
 
 #### `engine_args.max_num_seqs`
 
-The maximum number of sequences for concurrent processing in this stage. For LLM stages, this controls the vLLM scheduler's maximum concurrent sequences. For all stage types, this also controls how many tasks can be batched together in the task processing loop.
+The maximum number of sequences for concurrent processing in this stage. For
+LLM stages, this controls the vLLM scheduler's maximum concurrent sequences.
+For diffusion stages, this controls scheduler wave capacity in both request
+batch mode and step batch mode. `max_num_seqs: 1` is the serial/conservative
+path; values above `1` allow compatible requests to batch when the selected
+diffusion pipeline supports that batching mode.
 
 Default: `1`
 

docs/contributing/README.md

@@ -71,7 +71,7 @@ For additional features and advanced configurations, refer to the:
 ### Testing
 
 vLLM-Omni uses `pytest` to test the codebase.
-Please refer to the [test instructions](./ci/test_guide.md) for detailed testing information.
+Please refer to the [test instructions](./ci/test_execution_guide.md) for detailed testing information.
 
 !!! warning
     Currently, not all unit tests pass when run on CPU platforms. If you don't have access to a GPU platform to run unit tests locally, rely on the continuous integration system to run the tests for now.
@@ -127,11 +127,11 @@ The skill offers two modes:
 The precheck covers five PR types: Bug Fix, Performance, New Model, Diffusion Model, and General. Each type has a tailored checklist that validates evidence quality (repro steps, A/B benchmarks, registry entries, etc.). See the [precheck-pr skill](https://github.com/vllm-project/vllm-omni/blob/main/.claude/skills/precheck-pr/SKILL.md) for the full checklist.
 
 ### Local Test
-Please run the L1 and L2 test cases locally first and attach the results before contacting us to add the "ready" label. Please refer to the [test instructions](./ci/test_guide.md) for running the test cases.
+Please run the L1 and L2 test cases locally first and attach the results before contacting us to add the "ready" label. Please refer to the [test instructions](./ci/test_execution_guide.md) for running the test cases.
 
 ### Automatic skip-ci (docs and pytest skip marks)
 
-On pull requests and `main` pushes, the bootstrap step in [`.buildkite/pipeline.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/pipeline.yml) runs [`.buildkite/scripts/upload_pipeline.py`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/scripts/upload_pipeline.py) against the git diff. When every changed file qualifies, **L2 (`ready`) and L3 (`merge-test`) pipelines are not uploaded**, so the default GPU CI jobs are skipped.
+On pull requests and `main` pushes, the bootstrap step in [`.buildkite/cuda/pipeline.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/pipeline.yml) runs [`.buildkite/common/scripts/upload_pipeline.py`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/common/scripts/upload_pipeline.py) against the git diff. When every changed file qualifies, **L2 (`ready`) and L3 (`merge-test`) pipelines are not uploaded**, so the default GPU CI jobs are skipped.
 
 | Change per file | Examples |
 | --- | --- |
@@ -148,7 +148,7 @@ These PR shapes all trigger skip-ci:
 Skip-ci does **not** apply when the diff also touches product code (for example `vllm_omni/`), or when test files change assertions, imports, fixtures, or other non-skip logic. If the diff cannot be resolved (non-PR branches outside `main`), CI runs as usual.
 
 !!! note
-    Skipping L2/L3 does **not** disable the Docker image build step. Nightly (L4) upload can still run when the PR has a `nightly-test` label or on scheduled `main` builds with `NIGHTLY=1`. See [CI levels](./ci/CI_5levels.md) for how bootstrap skip-ci relates to diff-gated E2E jobs.
+    Skipping L2/L3 does **not** disable the Docker image build step. Nightly (L4) upload can still run when the PR has a `nightly-test` label or on scheduled `main` builds with `NIGHTLY=1`. Bootstrap child steps live in `bootstrap-upload-steps.yml`; the hook entry step runs `upload_pipeline.py --upload <platform>/bootstrap-upload-steps.yml`, which injects `if` by step `key` from skip-ci before upload. See [CI Settings — Diff-aware CI](./ci/ci_settings.md#diff-aware-ci) and [Test System Overview](./ci/test_system_overview.md).
 
 ### Code Quality
 

docs/contributing/ci/.nav.yaml

@@ -1,6 +1,6 @@
 nav:
-  - CI_5levels.md
+  - test_system_overview.md
+  - ci_settings.md
+  - test_writing_guide.md
   - failures.md
-  - test_guide.md
-  - test_markers.md
-  - test_style.md
+  - test_execution_guide.md

docs/contributing/ci/ci_settings.md

@@ -0,0 +1,344 @@
+# CI Settings
+
+This document describes **where** Buildkite YAML lives in the repo, **how each platform organizes CI**, and **how to add a new job**. It does not document agent queues, GPU types, or container plugin details—those belong in infra / preset files (for example `.buildkite/common/ci_mirror_hardwares.yml`).
+
+For CI levels (L1–L5) and triggers, see [Test System Overview](./test_system_overview.md). For test authoring, see [Test Writing Guide](./test_writing_guide.md). For running tests locally or replaying CI jobs, see [Test Execution Guide](./test_execution_guide.md).
+
+## Directory layout
+
+Canonical layout (prefer these paths for new changes):
+
+```
+.buildkite/
+├── common/                          # Shared across platforms
+│   ├── scripts/
+│   │   ├── skip_ci.py               # skip-ci decision (docs / skip-mark / CI YAML paths)
+│   │   ├── upload_pipeline.py       # Bootstrap + test-pipeline uploader (CUDA/NPU)
+│   │   └── resolve_skip_ci.sh       # Shell helpers for AMD/Intel bootstrap
+│   └── ci_mirror_hardwares.yml      # CUDA uploader presets (referenced by name only)
+├── cuda/                            # Primary NVIDIA CUDA CI
+│   ├── pipeline.yml                 # Bootstrap entry (hook upload)
+│   ├── bootstrap-upload-steps.yml   # Bootstrap child steps (upload_pipeline --upload)
+│   ├── test-ready.yml               # L2
+│   ├── test-merge.yml               # L3
+│   ├── test-nightly.yml             # L4
+│   ├── test-weekly.yml              # L5
+│   └── rebase-pipeline.yml
+├── npu/
+│   ├── pipeline-npu.yml             # Bootstrap entry (hook upload)
+│   ├── bootstrap-upload-steps.yml   # Bootstrap child steps (upload_pipeline --upload)
+│   ├── pipeline-npu-a3.yml          # A3 variant (when used)
+│   ├── test-npu-ready.yml           # L2
+│   ├── test-npu-nightly.yml         # L4
+│   └── scripts/
+├── amd/
+│   ├── test-amd-ready.yml           # L2 job definitions (template input)
+│   ├── test-amd-merge.yml           # L3 job definitions
+│   ├── test-template-amd-omni.j2    # Renders final pipeline.yaml
+│   └── scripts/
+│       ├── bootstrap-amd-omni.sh    # Entry: skip-ci → Jinja → upload
+│       └── run-amd-test.sh          # Wraps pytest inside ROCm docker
+├── intel/
+│   ├── pipeline-intel.yml           # Static Intel XPU pipeline
+│   └── scripts/
+│       ├── bootstrap-intel-omni.sh
+│       └── run-xpu-test.sh
+└── release/
+    ├── release-pipeline.yml
+    └── scripts/
+```
+
+**Placement rules**
+
+| Rule | Detail |
+| ---- | ------ |
+| Platform code under platform dir | New CUDA jobs go in `.buildkite/cuda/`; do not add new top-level `.buildkite/test-*.yml` files. |
+| Shared logic in `common/` | Skip-ci and CUDA upload rendering stay in `.buildkite/common/scripts/`. |
+| Bootstrap vs test YAML | **Bootstrap** (`pipeline*.yml`) builds images and uploads **child** test pipelines. **Test** YAML (`test-*.yml`) lists pytest steps only. |
+| Register CI YAML in `skip_ci.py` | If you add a new whitelisted test pipeline file, update `L2_YAML_FILES`, `L3_YAML_FILES`, or `L45_YAML_FILES` in `.buildkite/common/scripts/skip_ci.py` so skip-ci paths stay correct. |
+
+There are still **legacy copies** at `.buildkite/*.yaml` (without the `cuda/` prefix). Treat `.buildkite/cuda/*` as source of truth.
+
+## Platform comparison
+
+| Platform | Bootstrap entry | Test job files | Upload mechanism | Job hardware in YAML |
+| -------- | ---------------- | -------------- | ---------------- | -------------------- |
+| **CUDA** | `cuda/pipeline.yml` | `test-ready.yml`, `test-merge.yml`, `test-nightly.yml`, `test-weekly.yml` | `upload_pipeline.py --upload` (expands uploader-only keys) | `mirror_hardwares: <preset>` (string) |
+| **NPU** | `npu/pipeline-npu.yml` | `test-npu-ready.yml`, `test-npu-nightly.yml` | `upload_pipeline.py --upload` | `mirror_hardwares: a2b3_npu_1` / `a2b3_npu_4` / `a3_npu_2` |
+| **AMD** | `amd/scripts/bootstrap-amd-omni.sh` | `test-amd-ready.yml`, `test-amd-merge.yml` | Jinja (`test-template-amd-omni.j2`) → `pipeline upload` | `agent_pool` + `mirror_hardwares: [amdproduction]` (array, template filter) |
+| **Intel** | `intel/scripts/bootstrap-intel-omni.sh` | `intel/pipeline-intel.yml` (steps inline) | Direct `pipeline upload` | Inline `agents.queue` on each step |
+
+## Platform configuration style
+
+=== "CUDA"
+
+    **Bootstrap:** [`cuda/pipeline.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/pipeline.yml) (hook upload) + [`cuda/bootstrap-upload-steps.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/bootstrap-upload-steps.yml) (child steps). Document 1 runs [`upload_pipeline.py --upload .buildkite/cuda/bootstrap-upload-steps.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/common/scripts/upload_pipeline.py), which injects `if` by step `key` from skip-ci and uploads image build plus L2–L5 child pipeline upload steps (see [Diff-aware CI — Bootstrap skip](#bootstrap-skip)).
+
+    **Test YAML:** `cuda/test-ready.yml` (L2), `test-merge.yml` (L3), `test-nightly.yml` (L4), `test-weekly.yml` (L5). Each file starts with shared `env:` then `steps:`.
+
+    | CI level | File | Typical trigger |
+    | -------- | ---- | ----------------- |
+    | L2 | `cuda/test-ready.yml` | `ready` label |
+    | L3 | `cuda/test-merge.yml` | `merge-test` label / main merge |
+    | L4 | `cuda/test-nightly.yml` | `nightly-test` label or `NIGHTLY=1` |
+    | L5 | `cuda/test-weekly.yml` | `weekly-test` label or `WEEKLY=1` |
+
+    **Upload:** `upload_pipeline.py --upload` expands uploader-only keys before Buildkite upload.
+
+    **Hardware in YAML:** `mirror_hardwares: <preset>` (string)—preset names in [`common/ci_mirror_hardwares.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/common/ci_mirror_hardwares.yml). Do **not** set `agents` / `plugins` on the same step.
+
+    **Conventions**
+
+    - **`depends_on`:** leaf jobs depend on `upload-ready-pipeline`, `upload-merge-pipeline`, etc.
+    - **`group` / `label`:** `:card_index_dividers:` groups; labels like `Diffusion · Qwen Image Test`.
+    - **`commands`:** `timeout … pytest …` with markers and `--run-level` for the pipeline level.
+    - **`source_file_dependencies`:** required on **E2E Test** leaf jobs in L2/L3; see [Step filtering](#step-filtering).
+
+    **Adding a job**
+
+    1. Pick the level file (ready / merge / nightly / weekly).
+    2. Add a step under the right **group** (usually **E2E Test** for model pytest).
+    3. Set `label`, `commands`, `mirror_hardwares`, `depends_on: upload-<level>-pipeline`.
+    4. For L2/L3 E2E, add `source_file_dependencies` (pytest + model + deploy YAML prefixes).
+    5. Dry-run:
+
+    ```bash
+    python3 .buildkite/common/scripts/upload_pipeline.py .buildkite/cuda/test-ready.yml
+    ```
+
+=== "NPU"
+
+    **Bootstrap:** [`npu/pipeline-npu.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/npu/pipeline-npu.yml) + [`npu/bootstrap-upload-steps.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/npu/bootstrap-upload-steps.yml)—same split as CUDA; builds A2/B3 and A3 CI images, then uploads child test pipelines.
+
+    **Test YAML:** `npu/test-npu-ready.yml` (L2), `test-npu-nightly.yml` (L4).
+
+    **Upload:** same as CUDA—`upload_pipeline.py --upload`.
+
+    **Hardware in YAML:** `mirror_hardwares` preset (string), expanded to `agents`, top-level `image`, and `plugins`. Presets: `a2b3_npu_1`, `a2b3_npu_4`, `a3_npu_2` in `common/ci_mirror_hardwares.yml`.
+
+    **Conventions**
+
+    - **`depends_on: upload-ready-pipeline`** (or `upload-nightly-pipeline`) ties jobs to bootstrap upload keys.
+    - Do not duplicate `agents` / `image` / `plugins` when using `mirror_hardwares`.
+
+    **Adding a job**
+
+    1. Edit `test-npu-ready.yml` (L2) or `test-npu-nightly.yml` (L4).
+    2. Add a step with `mirror_hardwares` (add a new preset in `ci_mirror_hardwares.yml` first if needed).
+    3. Set `commands` to your pytest file and markers.
+    4. Dry-run:
+
+    ```bash
+    python3 .buildkite/common/scripts/upload_pipeline.py .buildkite/npu/test-npu-ready.yml
+    ```
+
+    5. Register new pipeline paths in `skip_ci.py` (`L2_YAML_FILES` or `L45_YAML_FILES`) when applicable.
+
+=== "AMD"
+
+    **Bootstrap:** [`amd/scripts/bootstrap-amd-omni.sh`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/amd/scripts/bootstrap-amd-omni.sh)—skip-ci, diff filtering, Jinja render, then `buildkite-agent pipeline upload`.
+
+    **Test YAML (data):** `amd/test-amd-ready.yml` (L2 / PR), `test-amd-merge.yml` (L3 / main).
+
+    **Rendering:** [`test-template-amd-omni.j2`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/amd/test-template-amd-omni.j2) wraps data steps with `amd-build` image build and `amd_<agent_pool>` queues. Do **not** hand-edit generated `pipeline.yaml`.
+
+    **Hardware in YAML:** `agent_pool` (for example `mi325_1`) plus `mirror_hardwares: [amdproduction]` (array—Buildkite template filter, not the CUDA/NPU uploader preset mechanism).
+
+    **Data file fields**
+
+    | Field | Purpose |
+    | ----- | ------- |
+    | `agent_pool` | ROCm pool; template maps to `queue: amd_<pool>`. |
+    | `mirror_hardwares` | Which mirror HW runs the step (for example `[amdproduction]`). |
+    | `commands` | Passed into `run-amd-test.sh` via `TEST_COMMAND`. |
+
+    **Adding a job**
+
+    1. Edit `test-amd-ready.yml` or `test-amd-merge.yml`.
+    2. Copy a neighboring block: `label`, `agent_pool`, `mirror_hardwares`, `commands`, optional `grade`.
+    3. Regenerate via bootstrap / Jinja; update `skip_ci.py` if you add a new YAML path.
+
+=== "Intel"
+
+    **Bootstrap:** [`intel/scripts/bootstrap-intel-omni.sh`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/intel/scripts/bootstrap-intel-omni.sh)—skip-ci, then direct `pipeline upload`.
+
+    **Test YAML:** steps live inline in [`intel/pipeline-intel.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/intel/pipeline-intel.yml); runners under `intel/scripts/` (for example `run-xpu-test.sh`).
+
+    **Upload:** `buildkite-agent pipeline upload` (no `upload_pipeline.py` mirror expansion).
+
+    **Hardware in YAML:** inline `agents.queue` (for example `intel-gpu-omni`) on each step.
+
+    **Adding a job**
+
+    1. Add a step to `pipeline-intel.yml`, or extend an existing runner script.
+    2. Match `agents`, `env`, `timeout_in_minutes`, and `command`/`commands` with sibling steps.
+    3. Keep `pipeline-intel.yml` listed in `L2_YAML_FILES` in `skip_ci.py`.
+
+## Cross-cutting conventions
+
+### Diff-aware CI {#diff-aware-ci}
+
+PR diffs drive **two independent skip layers**. Both read changed files from git, but at different pipeline stages and with different granularity:
+
+| Layer | Script | When | What is skipped | Where you configure |
+| ----- | ------ | ---- | --------------- | ------------------- |
+| **Bootstrap** | [`skip_ci.py`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/common/scripts/skip_ci.py) + [`upload_pipeline.py`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/common/scripts/upload_pipeline.py) | Before child test pipelines upload (`cuda/pipeline.yml`, `npu/pipeline-npu.yml`, AMD/Intel bootstraps) | Entire default CI, or whole L2/L3 upload for a platform | Whitelists in `skip_ci.py`; bootstrap `if` injected by step `key` in `upload_pipeline.py` |
+| **Step filter** | [`upload_pipeline.py`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/common/scripts/upload_pipeline.py) | While uploading CUDA L2/L3 YAML | Individual Buildkite steps inside `test-ready.yml` / `test-merge.yml` | `source_file_dependencies` on each step or group |
+
+**Changed files** (both layers):
+
+| Build context | Diff command |
+| --- | --- |
+| Pull request | `git diff --name-only origin/<base>...<BUILDKITE_COMMIT>` |
+| `main` push | `git diff --name-only <commit>^..<commit>` |
+| Local dry-run / non-PR | Diff unavailable → bootstrap and step filter both keep all steps; uploader-only keys are still stripped |
+
+Label triggers (`ready`, `merge-test`) are unchanged—diff-aware logic only reduces what runs **after** a pipeline is already scheduled.
+
+#### Bootstrap skip {#bootstrap-skip}
+
+[`skip_ci.py`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/common/scripts/skip_ci.py) classifies the git diff into buckets (docs / qualifying skip-mark / whitelisted L2·L3·L4/L5 YAML / other), then picks one decision path. CUDA/NPU apply that decision by injecting Buildkite `if` on `bootstrap-upload-steps.yml` step keys; AMD/Intel call `gate_bootstrap_ci <platform> <l2|l3>`.
+
+##### Decision overview
+
+| Diff shape | Path | Default L2/L3 |
+| --- | --- | --- |
+| Product code / non-whitelisted paths | normal CI | all on |
+| Docs and/or qualifying skip-mark only | `skip_all` | all off (scheduled `main` NIGHTLY/WEEKLY exceptions below) |
+| Whitelisted CI YAML only | yaml-gated (`skip_l2_l3`) | per platform/level matrix |
+| Docs/skip-mark **+** whitelisted CI YAML | yaml-gated (`skip_l2_l3`) | same as CI-YAML-only (does **not** widen to normal CI) |
+| Non-qualifying skip-mark and no CI YAML | normal CI | all on |
+| Diff unavailable | normal CI | all on |
+
+**`skip_all` exceptions (CUDA/NPU bootstrap only):** PR labels (`nightly-test`, `merge-test`, `npu-test`, `weekly-test`, …) do **not** revive jobs. On `main`, scheduled `NIGHTLY=1` still builds the image and uploads L4 plus L2/L3 (with `--e2e`); `WEEKLY=1` still builds the image and uploads L5 (CUDA).
+
+**Yaml-gated nightly/weekly:** L4/L5 upload steps keep their normal label / `NIGHTLY` / `WEEKLY` conditions (for example PR `nightly-test` still uploads nightly). Only L2/L3 upload steps are matrix-gated.
+
+##### Whitelisted CI YAML → platform
+
+Register new files in `L2_YAML_FILES`, `L3_YAML_FILES`, or `L45_YAML_FILES` in `skip_ci.py`.
+
+| Level | File | Platform |
+| --- | --- | --- |
+| L2 | `.buildkite/cuda/test-ready.yml` | cuda |
+| L2 | `.buildkite/npu/test-npu-ready.yml` | npu |
+| L2 | `.buildkite/amd/test-amd-ready.yml` | amd |
+| L2 | `.buildkite/intel/pipeline-intel.yml` | intel |
+| L3 | `.buildkite/cuda/test-merge.yml` | cuda |
+| L3 | `.buildkite/amd/test-amd-merge.yml` | amd |
+| L4/L5 | `.buildkite/cuda/test-nightly.yml`, `test-weekly.yml` | cuda |
+| L4/L5 | `.buildkite/npu/test-npu-nightly.yml` | npu |
+
+##### Yaml-gated category branches
+
+`L2?` / `L3?` / `L45?` = whether any file in that whitelist changed. Untouched platforms/levels end up off.
+
+| ID | L2? | L3? | L45? | Enabled L2/L3 |
+| --- | --- | --- | --- | --- |
+| G1 | | | ✓ | none (all L2/L3 off) |
+| G2 | ✓ | | | L2 only on platforms whose L2 YAML changed |
+| G3 | | ✓ | | L3 only on platforms whose L3 YAML changed |
+| G4 | ✓ | ✓ | | union of G2 and G3 (same platform ready+merge → both on) |
+| G5 | ✓ | | ✓ | same as G2 (L45 rescued by L2 YAML) |
+| G6 | | ✓ | ✓ | same as G3 (L45 rescued by L3 YAML) |
+| G7 | ✓ | ✓ | ✓ | same as G4 |
+
+##### Yaml-gated examples (platform-level)
+
+Docs/skip-mark mixed with any row below follows the same matrix.
+
+| Diff (whitelist) | Branch | Enabled L2/L3 |
+| --- | --- | --- |
+| Only `cuda/test-ready.yml` | G2 | `cuda/l2` |
+| Only `amd/test-amd-ready.yml` | G2 | `amd/l2` |
+| Only `npu/test-npu-ready.yml` | G2 | `npu/l2` |
+| Only `intel/pipeline-intel.yml` | G2 | `intel/l2` |
+| Only `cuda/test-merge.yml` | G3 | `cuda/l3` |
+| Only `amd/test-amd-merge.yml` | G3 | `amd/l3` |
+| CUDA ready + CUDA merge | G4 | `cuda/l2` + `cuda/l3` |
+| CUDA ready + AMD merge | G4 | `cuda/l2` + `amd/l3` |
+| CUDA ready + AMD ready | G2 | `cuda/l2` + `amd/l2` |
+| CUDA merge + AMD merge | G3 | `cuda/l3` + `amd/l3` |
+| Only nightly / weekly / npu-nightly YAML | G1 | (none) |
+| CUDA ready + nightly | G5 | `cuda/l2` |
+| CUDA merge + nightly | G6 | `cuda/l3` |
+| CUDA ready + merge + nightly | G7 | `cuda/l2` + `cuda/l3` |
+
+##### How platforms consume the decision
+
+| Platform | Mechanism |
+| --- | --- |
+| **CUDA** | `upload_pipeline.py` injects `if` by step `key` (`image-build`, `upload-ready-pipeline`, `upload-merge-pipeline`, …) from `is_run(cuda, l2/l3)` |
+| **NPU** | Same injection; no merge upload step (L3 always off in bootstrap) |
+| **AMD / Intel** | `gate_bootstrap_ci <platform> <l2\|l3>` exits 0 when `skip_all` or that target is off |
+
+Unit coverage: `tests/buildkite/test_skip_ci.py`.
+
+#### Step filtering {#step-filtering}
+
+CUDA **L2** (`.buildkite/cuda/test-ready.yml`) and **L3** (`.buildkite/cuda/test-merge.yml`) only. Bootstrap upload entry: `upload_pipeline.py --upload .buildkite/cuda/bootstrap-upload-steps.yml`.
+
+**Uploader-only keys** — removed before Buildkite sees the YAML; never used at runtime on agents:
+
+| Key | Purpose |
+| --- | ------- |
+| `source_file_dependencies` | List of path **prefixes**. If any changed file equals a prefix or starts with `prefix/`, keep the step (or group); otherwise omit it. |
+| `mirror_hardwares` | Expand to `agents` + `plugins` (+ optional `image`) from `ci_mirror_hardwares.yml`. |
+
+**Policy**
+
+- **Always uploaded** (no key): groups outside **E2E Test**—Simple Test, Diffusion unit tests, Engine/Model Executor, Distributed, Custom Pipeline, Entrypoints (L2), LoRA / Entrypoints (L3).
+- **Diff-gated**: every **E2E Test** leaf job. List the smallest prefix set per step—pytest file(s), model code under `vllm_omni/model_executor/models/` or `vllm_omni/diffusion/models/`, plus `stage_input_processors/` and `vllm_omni/deploy/*.yaml` when applicable. A **group** may define the key instead; the whole group drops if no prefix matches.
+
+**YAML examples**
+
+```yaml
+      - label: "Diffusion · Qwen Image Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen_image.py
+          - vllm_omni/diffusion/models/qwen_image/
+        commands:
+          - pytest -s -v tests/e2e/online_serving/test_qwen_image.py -m 'core_model' ...
+        mirror_hardwares: h100_1
+
+      - label: "TTS · Qwen3-TTS CustomVoice Test"
+        source_file_dependencies:
+          - tests/e2e/online_serving/test_qwen3_tts_customvoice.py
+          - vllm_omni/model_executor/models/qwen3_tts/
+          - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
+          - vllm_omni/deploy/qwen3_tts.yaml
+        commands:
+          - pytest -s -v tests/e2e/online_serving/test_qwen3_tts_customvoice.py ...
+        mirror_hardwares: l4_4
+```
+
+**Local dry-run**
+
+```bash
+python3 .buildkite/common/scripts/upload_pipeline.py .buildkite/cuda/test-ready.yml
+python3 .buildkite/common/scripts/upload_pipeline.py .buildkite/cuda/test-merge.yml | grep source_file_dependencies
+# (no output expected)
+```
+
+On PR builds, `upload_pipeline.py` logs `skip '…' (no changes under …)` for omitted steps.
+
+### Labels and grouping
+
+- Use **groups** for dashboard readability; keep **E2E Test** as the group name CUDA diff filtering expects for `--e2e` nightly runs on main.
+- Prefix labels by model domain: **Omni ·**, **TTS ·**, **Diffusion ·**, **Simple ·**, etc., matching existing steps.
+
+### Validation checklist
+
+| Check | Command / location |
+| ----- | ------------------ |
+| CUDA render | `python3 .buildkite/common/scripts/upload_pipeline.py .buildkite/cuda/test-<level>.yml` |
+| No leaked uploader keys | `… \| grep -E 'mirror_hardwares|source_file_dependencies'` → empty |
+| Skip-ci / upload unit tests | `pytest tests/buildkite/` |
+| Local job replay (CUDA L2+) | `tools/run_ready_jobs.sh`, `tools/run_merge_jobs.sh`, `tools/nightly/run_nightly_jobs.sh` (read YAML from `cuda/`) |
+
+## Related documentation
+
+- [Test System Overview](./test_system_overview.md) — L1–L5 scope, triggers, test pyramid
+- [Test Writing Guide](./test_writing_guide.md) — markers, directories, L1–L5 examples
+- [Test Execution Guide](./test_execution_guide.md) — running CI-aligned jobs locally
+- Implementation: `.buildkite/common/scripts/upload_pipeline.py`, `.buildkite/common/scripts/skip_ci.py`

docs/contributing/ci/failures.md

@@ -79,7 +79,7 @@ For resilience behavior under injected process signals/OOM pressure (validation
 
 ### Running Tests Locally
 
-For instructions on running tests locally (including specific test files, functions, and markers), see the [Running Tests](https://docs.vllm.ai/projects/vllm-omni/en/latest/contributing/ci/test_guide/#running-tests) section in the Test Guide.
+For instructions on running tests locally (including specific test files, functions, and markers), see the [Running Tests](https://docs.vllm.ai/projects/vllm-omni/en/latest/contributing/ci/test_execution_guide/#running-tests) section in the Test Execution guide.
 
 ## Step 4: Raise an Issue or Fix It
 

docs/contributing/ci/test_examples/l4_functionality_tests.inc.md

@@ -37,7 +37,7 @@ Currently all the features are available in online serving mode. Hence, only nee
 **Code Style**
 
 - Validation: test that the multimodal output files of your model have the correct shapes. `OpenAIClientHandler.send_diffusion_request` should have taken care of this.
-- Test marks: always add `full_model` and `diffusion` for L4 nightly `test_*_expansion.py` cases. Add GPU-related marks if needed. Ref: [Markers for Tests](https://docs.vllm.ai/projects/vllm-omni/en/latest/contributing/ci/tests_markers/).
+- Test marks: always add `full_model` and `diffusion` for L4 nightly `test_*_expansion.py` cases. Add GPU-related marks if needed. Ref: [Markers for Tests](https://docs.vllm.ai/projects/vllm-omni/en/latest/contributing/ci/test_writing_guide/#markers-for-tests).
 - To maximize code reuse, you may refer to
     - `tests/conftest.py` for `omni_server` (running server in subprocess) and `openai_client` fixtures (sending requests and validating output), `generate_synthetic_image` and `assert_XXX_valid` helper.
     - `tests/helpers/mark.py` for `@hardware_test(...)` and `hardware_marks`.

docs/contributing/ci/test_examples/l4_performance_tests.inc.md

@@ -148,7 +148,7 @@ pytest -sv tests/dfx/perf/scripts/run_diffusion_benchmark.py --assert-baseline \
   --test-config-file tests/dfx/perf/tests/test_qwen_image_vllm_omni.json
 ```
 
-See also [Markers for Tests](./tests_markers.md) for registered hardware markers (`H100`, `L4`, `cuda`, `distributed_cuda`, …).
+See also [Markers for Tests](#markers-for-tests) for registered hardware markers (`H100`, `L4`, `cuda`, `distributed_cuda`, …).
 
 **`server_params` Configuration**
 

docs/contributing/ci/test_execution_guide.md

@@ -1,4 +1,4 @@
-# Test Guide
+# Test Execution Guide
 ## Setting Up the Test Environment
 ### Creating a Container
 vLLM-Omni provides an official Docker image for deployment. These images are built upon vLLM Docker images and are available on [Docker Hub](https://hub.docker.com/r/vllm/vllm-omni/tags). The version of vLLM-Omni indicates which vLLM release it is based on.
@@ -49,11 +49,11 @@ Failed jobs: 1/2
     cd tests
     pytest -s -v -m "core_model and cpu"
     ```
-    The latest test command is available in the "Simple Unit Test" step of this [pipeline](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/test-ready.yml).
+    The latest test command is available in the "Simple Unit Test" step of this [pipeline](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/test-ready.yml).
 
 === "L2 level"
 
-    **Recommended:** run CI-aligned jobs from the repo root via [`tools/run_ready_jobs.sh`](https://github.com/vllm-project/vllm-omni/blob/main/tools/run_ready_jobs.sh). The script reads [`.buildkite/test-ready.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/test-ready.yml), generates per-step bash wrappers, runs pytest, and tees logs under `logs/ready_jobs/` (see [CI job runners](#ci-job-runners-l2l4-logs-timing-and-timeouts) for timing and timeout behavior). Requires `bash`, `python3`, and PyYAML (`pip install pyyaml`).
+    **Recommended:** run CI-aligned jobs from the repo root via [`tools/run_ready_jobs.sh`](https://github.com/vllm-project/vllm-omni/blob/main/tools/run_ready_jobs.sh). The script reads [`.buildkite/cuda/test-ready.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/test-ready.yml), generates per-step bash wrappers, runs pytest, and tees logs under `logs/ready_jobs/` (see [CI job runners](#ci-job-runners-l2l4-logs-timing-and-timeouts) for timing and timeout behavior). Requires `bash`, `python3`, and PyYAML (`pip install pyyaml`).
 
     ```bash
     # All L2 and L1 jobs (default: every pytest step in test-ready.yml)
@@ -85,7 +85,7 @@ Failed jobs: 1/2
 
 === "L3 level"
 
-    **Recommended:** run CI-aligned jobs from the repo root via [`tools/run_merge_jobs.sh`](https://github.com/vllm-project/vllm-omni/blob/main/tools/run_merge_jobs.sh). The script reads [`.buildkite/test-merge.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/test-merge.yml), generates per-step bash wrappers, runs pytest, and tees logs under `logs/merge_jobs/` (see [CI job runners](#ci-job-runners-l2l4-logs-timing-and-timeouts)). Requires `bash`, `python3`, and PyYAML (`pip install pyyaml`).
+    **Recommended:** run CI-aligned jobs from the repo root via [`tools/run_merge_jobs.sh`](https://github.com/vllm-project/vllm-omni/blob/main/tools/run_merge_jobs.sh). The script reads [`.buildkite/cuda/test-merge.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/test-merge.yml), generates per-step bash wrappers, runs pytest, and tees logs under `logs/merge_jobs/` (see [CI job runners](#ci-job-runners-l2l4-logs-timing-and-timeouts)). Requires `bash`, `python3`, and PyYAML (`pip install pyyaml`).
 
     ```bash
     # All L3 and L1 jobs (default: every pytest step in test-merge.yml)
@@ -117,7 +117,7 @@ Failed jobs: 1/2
 
 === "L4 level"
 
-    **Recommended:** run CI-aligned nightly jobs from the repo root via [`tools/nightly/run_nightly_jobs.sh`](https://github.com/vllm-project/vllm-omni/blob/main/tools/nightly/run_nightly_jobs.sh). The script reads [`.buildkite/test-nightly.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/test-nightly.yml), generates per-step bash wrappers, runs pytest (perf jobs first, then aggregates perf JSON into Excel), and tees logs under `logs/nightly_jobs/` (see [CI job runners](#ci-job-runners-l2l4-logs-timing-and-timeouts); `generate_nightly_perf_excel` is included in the timing summary). Requires `bash`, `python3`, and PyYAML (`pip install pyyaml`).
+    **Recommended:** run CI-aligned nightly jobs from the repo root via [`tools/nightly/run_nightly_jobs.sh`](https://github.com/vllm-project/vllm-omni/blob/main/tools/nightly/run_nightly_jobs.sh). The script reads [`.buildkite/cuda/test-nightly.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/test-nightly.yml), generates per-step bash wrappers, runs pytest (perf jobs first, then aggregates perf JSON into Excel), and tees logs under `logs/nightly_jobs/` (see [CI job runners](#ci-job-runners-l2l4-logs-timing-and-timeouts); `generate_nightly_perf_excel` is included in the timing summary). Requires `bash`, `python3`, and PyYAML (`pip install pyyaml`).
 
     ```bash
     # All nightly jobs (default: test-type all, model-type all)
@@ -189,9 +189,9 @@ Failed jobs: 1/2
 
     ```
 
-    The latest L5 CI jobs (reliability + invalid-parameter weekly steps) are in [test-weekly.yml](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/test-weekly.yml).
+    The latest L5 CI jobs (reliability + invalid-parameter weekly steps) are in [test-weekly.yml](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/test-weekly.yml).
 
-You can find more information about markers in the documentation: [marker doc](./tests_markers.md)
+You can find more information about markers in [Test Writing Guide — Markers for Tests](./test_writing_guide.md#markers-for-tests).
 
 ## Adding New Test Cases
-Please refer to the [L5 Layering Specification document](./CI_5levels.md).
+Please refer to the [Test System Overview](./test_system_overview.md) and [Test Writing Guide](./test_writing_guide.md).

docs/contributing/ci/test_system_overview.md

@@ -0,0 +1,210 @@
+# Multi-Level Automated Testing System Overview
+
+## Document Overview
+
+This testing system aims to build a complete, efficient, and well-structured quality assurance framework for the development, integration, and release of model services. It draws on the concept of the test pyramid from modern software engineering, progressively expanding testing activities from basic code logic verification to complex end-to-end (E2E) functionality, performance, accuracy, and even long-term stability validation.
+
+Through five levels (L1-L5) and common (Common) specifications, the system clarifies the testing objectives, scope, execution frequency, and required resources for different development stages (e.g., each commit, PR merge, daily build, pre-release). This ensures that models meet high standards for functionality, performance, and reliability across various deployment scenarios (online serving and offline inference).
+
+<table>
+  <thead>
+    <tr>
+      <th>Level</th>
+      <th>Scope & Focus</th>
+      <th>Model Coverage Strategy</th>
+      <th>Feature Coverage Strategy</th>
+      <th>Interface Coverage Strategy</th>
+      <th>Tags</th>
+      <th>Time Cost</th>
+      <th>Test Dir</th>
+      <th>Doc</th>
+      <th>Frequency</th>
+      <th>Hardware</th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td rowspan="2"><strong>Common</strong></td>
+      <td>Contribution Guideline & PR checklist</td>
+      <td>/</td>
+      <td>/</td>
+      <td>/</td>
+      <td>/</td>
+      <td>/</td>
+      <td>/</td>
+      <td><a href="https://github.com/vllm-project/vllm-omni/blob/main/.github/PULL_REQUEST_TEMPLATE.md">PR Checklist</a></td>
+      <td>/</td>
+      <td>/</td>
+    </tr>
+    <tr>
+      <td>CI Failure Description</td>
+      <td>/</td>
+      <td>/</td>
+      <td>/</td>
+      <td>/</td>
+      <td>/</td>
+      <td>/</td>
+      <td><a href="./failures.md">CI Failures</a></td>
+      <td>/</td>
+      <td>/</td>
+    </tr>
+    <tr>
+      <td><strong>L1</strong><br>(Unit & Logic)</td>
+      <td>Unit tests for components like entrypoints, models</td>
+      <td>/</td>
+      <td>/</td>
+      <td>/</td>
+      <td><code>core_model and cpu</code></td>
+      <td rowspan="2">&lt;15min</td>
+      <td>/tests/{component_name}/test_xxx</td>
+      <td>
+        <a href="test_writing_guide.md#l1--l2-level-testing-unit-testing-and-basic-end-to-end-verification">L1 &amp; L2</a><br>
+        Section 1 L1&amp;L2: Purpose, Test Content, Directory Location, Example
+      </td>
+      <td>PR with ready label (also can run locally)</td>
+      <td>CPU</td>
+    </tr>
+    <tr>
+      <td><strong>L2</strong><br>(E2E across models & GPU-required UT)</td>
+      <td>Online (basic deployment scenarios):<br>dummy, normal inference function (output format, stream), some instance startup UT</td>
+      <td>High-priority models + online basic scenarios; request success, non-empty output, format match (no Whisper/accuracy)</td>
+      <td>High-priority features (using random lightweight models)</td>
+      <td>High-priority interfaces (using random lightweight models)</td>
+      <td><code>core_model and hardware_test(H100, L4, etc.) and omni/tts/diffusion</code></td>
+      <td>
+        <strong>Model tests:</strong><br>
+        /tests/e2e/online_serving/test_{model_name}.py<br>
+        <strong>Feature tests:</strong><br>
+        /tests/{component_name}/test_xxx<br>
+        <strong>Interface tests:</strong><br>
+        /tests/entrypoints/test_xxx
+      </td>
+      <td>
+        <a href="test_writing_guide.md#l1--l2-level-testing-unit-testing-and-basic-end-to-end-verification">L1 &amp; L2</a><br>
+        L1&amp;L2: Purpose, Test Content, Directory Location, Example
+      </td>
+      <td>PR with ready label</td>
+      <td>GPU</td>
+    </tr>
+    <tr>
+      <td><strong>L3</strong><br>(Important Perf & Integration & Accuracy)</td>
+      <td>Online & Offline (multiple deployment scenarios):<br>real model, normal inference function, normal accuracy</td>
+      <td>High/medium-priority models + key online/offline scenarios; real weights, Whisper/similarity, preset voice gender, basic accuracy</td>
+      <td>Medium-priority features (using random lightweight models)</td>
+      <td>Medium-priority interfaces (using random lightweight models)</td>
+      <td><code>advanced_model and hardware_test(H100, L4, etc.) and omni/tts/diffusion</code></td>
+      <td>&lt;30min</td>
+      <td>
+        <strong>Model tests:</strong><br>
+        /tests/e2e/online_serving/test_{model_name}.py<br>
+        /tests/e2e/offline_inference/test_{model_name}.py<br>
+        <strong>Feature tests:</strong><br>
+        /tests/{component_name}/test_xxx<br>
+        <strong>Interface tests:</strong><br>
+        /tests/entrypoints/test_xxx
+      </td>
+      <td>
+        <a href="test_writing_guide.md#l3-level-testing-core-integration-performance-and-accuracy-verification">L3</a><br>
+        L3: Purpose, Test Content, Directory Location, Example
+      </td>
+      <td>PR Merged (Also run L1&L2 Tests)</td>
+      <td>GPU</td>
+    </tr>
+    <tr>
+      <td><strong>L4</strong><br>(Perf & Integration & Accuracy)</td>
+      <td>Online: full functional scenarios + performance test + doc test + accuracy test</td>
+      <td>High-priority models: function, performance, accuracy, and doc testing<br>Medium-priority models: function and doc testing</td>
+      <td>Low-priority features (using real weights)</td>
+      <td>Low-priority interfaces (using real weights)</td>
+      <td><code>full_model and hardware_test(H100, L4, etc.) and omni/tts/diffusion</code></td>
+      <td>&lt;3 hour</td>
+      <td>
+        <strong>Model tests:</strong><br>
+        /tests/e2e/online_serving/test_{model_name}_expansion.py<br>
+        <strong>Feature tests:</strong><br>
+        /tests/{component_name}/test_xxx<br>
+        <strong>Interface tests:</strong><br>
+        /tests/entrypoints/test_xxx<br>
+        <strong>Performance:</strong><br>
+        /tests/dfx/perf/tests/test_qwen3_omni_*.json (Omni), test_tts.json (TTS),<br>
+        test_voxcpm2.json, test_higgs_audio_v3.json, and<br>
+        /tests/dfx/perf/tests/test_{diffusion_model}_vllm_omni.json (Diffusion)<br>
+        <strong>Doc Test:</strong><br>
+        tests/examples/online_serving/test_{model_name}.py<br>
+        tests/examples/offline_inference/test_{model_name}.py<br>
+        <strong>Accuracy Test:</strong><br>
+        /tests/e2e/accuracy/test_{model_name}.py
+      </td>
+      <td>
+        <a href="test_writing_guide.md#l4-level-testing-full-functionality-performance-and-documentation-testing">L4</a><br>
+        L4: Purpose, Test Content, Directory Location, Example
+      </td>
+      <td>Nightly</td>
+      <td>GPU</td>
+    </tr>
+    <tr>
+      <td><strong>L5</strong><br>(Stability & Reliability)</td>
+      <td>Online: long-term stability test + reliability test</td>
+      <td>Long-term stability and reliability testing for high-priority models<br>Low-priority models: function and doc testing</td>
+      <td>/</td>
+      <td>Invalid-parameter validation for high-priority interfaces</td>
+      <td><code>slow and hardware_test(H100, L4, etc.) and omni/tts/diffusion</code></td>
+      <td> Depends on reality </td>
+      <td>
+        <strong>Stability:</strong><br>
+        /tests/dfx/stability/tests/test_qwen3_omni.json<br>
+        /tests/dfx/stability/tests/test_wan22.json<br>
+        <strong>Reliability:</strong><br>
+        tests/dfx/reliability/test_reliability_{model_key}.py<br>
+        (e.g. <code>test_reliability_qwen3_omni.py</code>, <code>test_reliability_wan22.py</code>, <code>test_reliability_hunyuan_image.py</code>, <code>test_reliability_voxcpm2.py</code>)
+      </td>
+      <td>
+        <a href="test_writing_guide.md#l5-level-testing-stability-and-reliability-testing">L5</a><br>
+        L5: Purpose, Test Content, Directory Location, Example
+      </td>
+      <td>Weekly / Days before Release</td>
+      <td>GPU</td>
+    </tr>
+  </tbody>
+</table>
+
+For per-level test authoring (directories, markers, examples), see [Test Writing Guide](./test_writing_guide.md).
+
+## Common Specifications
+
+Before entering specific testing levels, the project establishes two common specifications aimed at standardizing the development process and quickly locating issues.
+
+1.  ***PR Checklist ([`.github/PULL_REQUEST_TEMPLATE.md`](https://github.com/vllm-project/vllm-omni/blob/main/.github/PULL_REQUEST_TEMPLATE.md))***: This template defines the self-check items that must be completed before submitting a code review (Pull Request). It ensures that each code change meets basic requirements such as code style, dependency updates, and documentation synchronization before entering the automated testing pipeline, serving as the first manual line of defense for quality assurance.
+2.  ***CI Failure Explanation ([CI Failures](./failures.md))***: This document archives and explains common failure patterns in the Continuous Integration (CI) pipeline, error log interpretation, and preliminary troubleshooting steps. It helps developers and testers quickly diagnose the causes of automated test failures, improving problem-solving efficiency.
+
+## Notes
+
+### L2 / L3 diff-aware CI (CUDA)
+
+On CUDA **L2** and **L3**, **E2E Test** Buildkite jobs may be omitted at pipeline upload when the PR diff does not touch their path prefixes; other groups still always upload. Full two-layer mechanics and YAML examples: [CI Settings — Diff-aware CI](./ci_settings.md#diff-aware-ci) ([step filtering](./ci_settings.md#step-filtering)).
+
+### Test helper environment variables
+
+Some shared helpers under `tests/helpers/` honor optional environment variables for local debugging. These are **not** set in CI by default.
+
+| Variable | Accepted values | Description |
+| -------- | --------------- | ----------- |
+| `VLLM_OMNI_KEEP_REQUEST_MEDIA` | `1`, `true`, `yes` (case-insensitive) | When enabled, temporary WAV files created by `tests.helpers.media.convert_audio_bytes_to_text` are **not** deleted when the pytest process exits. By default, each call writes a unique file under the system temp directory via `tempfile.mkstemp` and registers `atexit` cleanup. Use this when debugging audio output validation (Whisper transcription, keyword checks, text–audio similarity). The saved path is logged as `audio data is saved: <path>`. |
+
+Example (Linux / macOS):
+
+```bash
+export VLLM_OMNI_KEEP_REQUEST_MEDIA=1
+pytest -s -v tests/e2e/online_serving/test_qwen3_omni.py -k test_mix_to_text_audio
+```
+
+Example (Windows PowerShell):
+
+```powershell
+$env:VLLM_OMNI_KEEP_REQUEST_MEDIA = "1"
+pytest -s -v tests/e2e/online_serving/test_qwen3_omni.py -k test_mix_to_text_audio
+```
+
+## Summary
+
+This multi-level testing system achieves continuous, progressive validation of model service quality by tightly integrating testing activities with the development workflow (commit, review, merge, release). From rapid unit testing to comprehensive end-to-end testing, and further to in-depth performance, stability, and reliability verification, each level has clear objectives, collectively building a robust quality protection net. By following this system, teams can deliver high-quality, highly reliable model services more efficiently.

docs/contributing/ci/test_writing_guide.md

@@ -1,177 +1,9 @@
-# Multi-Level Automated Testing System Documentation
+# Test Writing Guide
 
-## Document Overview
+This document describes how to author tests for each CI level (L1–L5). For the level matrix, CI triggers, and Buildkite upload behavior, see [Test System Overview](./test_system_overview.md).
 
-This testing system aims to build a complete, efficient, and well-structured quality assurance framework for the development, integration, and release of model services. It draws on the concept of the test pyramid from modern software engineering, progressively expanding testing activities from basic code logic verification to complex end-to-end (E2E) functionality, performance, accuracy, and even long-term stability validation.
+## Test Directory Structure
 
-Through five levels (L1-L5) and common (Common) specifications, the system clarifies the testing objectives, scope, execution frequency, and required resources for different development stages (e.g., each commit, PR merge, daily build, pre-release). This ensures that models meet high standards for functionality, performance, and reliability across various deployment scenarios (online serving and offline inference).
-
-<table>
-  <thead>
-    <tr>
-      <th>Level</th>
-      <th>Scope & Focus</th>
-      <th>Model Coverage Strategy</th>
-      <th>Feature Coverage Strategy</th>
-      <th>Interface Coverage Strategy</th>
-      <th>Tags</th>
-      <th>Time Cost</th>
-      <th>Test Dir</th>
-      <th>Doc</th>
-      <th>Frequency</th>
-      <th>Hardware</th>
-    </tr>
-  </thead>
-  <tbody>
-    <tr>
-      <td rowspan="2"><strong>Common</strong></td>
-      <td>Contribution Guideline & PR checklist</td>
-      <td>/</td>
-      <td>/</td>
-      <td>/</td>
-      <td>/</td>
-      <td>/</td>
-      <td>/</td>
-      <td>.github/PULL_REQUEST_TEMPLATE.md <a href="../tests_style/"> Test Style (PR Checklist)</a></td>
-      <td>/</td>
-      <td>/</td>
-    </tr>
-    <tr>
-      <td>CI Failure Description</td>
-      <td>/</td>
-      <td>/</td>
-      <td>/</td>
-      <td>/</td>
-      <td>/</td>
-      <td>/</td>
-      <td><a href="../failures/"> CI Failures</a></td>
-      <td>/</td>
-      <td>/</td>
-    </tr>
-    <tr>
-      <td><strong>L1</strong><br>(Unit & Logic)</td>
-      <td>Unit tests for components like entrypoints, models</td>
-      <td>/</td>
-      <td>/</td>
-      <td>/</td>
-      <td><code>core_model and cpu</code></td>
-      <td rowspan="2">&lt;15min</td>
-      <td>/tests/{component_name}/test_xxx</td>
-      <td>
-        <a href="#chapter-1-l1-l2-level-testing-unit-testing-and-basic-end-to-end-verification">Chapter 1</a><br>
-        Section 1 L1&amp;L2: Purpose, Test Content, Directory Location, Example
-      </td>
-      <td>PR with ready label (also can run locally)</td>
-      <td>CPU</td>
-    </tr>
-    <tr>
-      <td><strong>L2</strong><br>(E2E across models & GPU-required UT)</td>
-      <td>Online (basic deployment scenarios):<br>dummy, normal inference function (output format, stream), some instance startup UT</td>
-      <td>High-priority models + online basic scenarios; request success, non-empty output, format match (no Whisper/accuracy)</td>
-      <td>High-priority features (using random lightweight models)</td>
-      <td>High-priority interfaces (using random lightweight models)</td>
-      <td><code>core_model and hardware_test(H100, L4, etc.) and omni/tts/diffusion</code></td>
-      <td>
-        <strong>Model tests:</strong><br>
-        /tests/e2e/online_serving/test_{model_name}.py<br>
-        <strong>Feature tests:</strong><br>
-        /tests/{component_name}/test_xxx<br>
-        <strong>Interface tests:</strong><br>
-        /tests/entrypoints/test_xxx
-      </td>
-      <td>
-        <a href="#chapter-1-l1-l2-level-testing-unit-testing-and-basic-end-to-end-verification">Chapter 1</a><br>
-        L1&amp;L2: Purpose, Test Content, Directory Location, Example
-      </td>
-      <td>PR with ready label</td>
-      <td>GPU</td>
-    </tr>
-    <tr>
-      <td><strong>L3</strong><br>(Important Perf & Integration & Accuracy)</td>
-      <td>Online & Offline (multiple deployment scenarios):<br>real model, normal inference function, normal accuracy</td>
-      <td>High/medium-priority models + key online/offline scenarios; real weights, Whisper/similarity, preset voice gender, basic accuracy</td>
-      <td>Medium-priority features (using random lightweight models)</td>
-      <td>Medium-priority interfaces (using random lightweight models)</td>
-      <td><code>advanced_model and hardware_test(H100, L4, etc.) and omni/tts/diffusion</code></td>
-      <td>&lt;30min</td>
-      <td>
-        <strong>Model tests:</strong><br>
-        /tests/e2e/online_serving/test_{model_name}.py<br>
-        /tests/e2e/offline_inference/test_{model_name}.py<br>
-        <strong>Feature tests:</strong><br>
-        /tests/{component_name}/test_xxx<br>
-        <strong>Interface tests:</strong><br>
-        /tests/entrypoints/test_xxx
-      </td>
-      <td>
-        <a href="#chapter-2-l3-level-testing-core-integration-performance-and-accuracy-verification">Chapter 2</a><br>
-        L3: Purpose, Test Content, Directory Location, Example
-      </td>
-      <td>PR Merged (Also run L1&L2 Tests)</td>
-      <td>GPU</td>
-    </tr>
-    <tr>
-      <td><strong>L4</strong><br>(Perf & Integration & Accuracy)</td>
-      <td>Online: full functional scenarios + performance test + doc test + accuracy test</td>
-      <td>High-priority models: function, performance, accuracy, and doc testing<br>Medium-priority models: function and doc testing</td>
-      <td>Low-priority features (using real weights)</td>
-      <td>Low-priority interfaces (using real weights)</td>
-      <td><code>full_model and hardware_test(H100, L4, etc.) and omni/tts/diffusion</code></td>
-      <td>&lt;3 hour</td>
-      <td>
-        <strong>Model tests:</strong><br>
-        /tests/e2e/online_serving/test_{model_name}_expansion.py<br>
-        <strong>Feature tests:</strong><br>
-        /tests/{component_name}/test_xxx<br>
-        <strong>Interface tests:</strong><br>
-        /tests/entrypoints/test_xxx<br>
-        <strong>Performance:</strong><br>
-        /tests/dfx/perf/tests/test_qwen3_omni_*.json (Omni), test_tts.json (TTS),<br>
-        test_voxcpm2.json, test_higgs_audio_v3.json, and<br>
-        /tests/dfx/perf/tests/test_{diffusion_model}_vllm_omni.json (Diffusion)<br>
-        <strong>Doc Test:</strong><br>
-        tests/examples/online_serving/test_{model_name}.py<br>
-        tests/examples/offline_inference/test_{model_name}.py<br>
-        <strong>Accuracy Test:</strong><br>
-        /tests/e2e/accuracy/test_{model_name}.py
-      </td>
-      <td>
-        <a href="#chapter-3-l4-level-testing-full-functionality-performance-and-documentation-testing">Chapter 3</a><br>
-        L4: Purpose, Test Content, Directory Location, Example
-      </td>
-      <td>Nightly</td>
-      <td>GPU</td>
-    </tr>
-    <tr>
-      <td><strong>L5</strong><br>(Stability & Reliability)</td>
-      <td>Online: long-term stability test + reliability test</td>
-      <td>Long-term stability and reliability testing for high-priority models<br>Low-priority models: function and doc testing</td>
-      <td>/</td>
-      <td>Invalid-parameter validation for high-priority interfaces</td>
-      <td><code>slow and hardware_test(H100, L4, etc.) and omni/tts/diffusion</code></td>
-      <td> Depends on reality </td>
-      <td>
-        <strong>Stability:</strong><br>
-        /tests/dfx/stability/tests/test_qwen3_omni.json<br>
-        /tests/dfx/stability/tests/test_wan22.json<br>
-        <strong>Reliability:</strong><br>
-        tests/dfx/reliability/test_reliability_{model_key}.py<br>
-        (e.g. <code>test_reliability_qwen3_omni.py</code>, <code>test_reliability_wan22.py</code>, <code>test_reliability_hunyuan_image.py</code>, <code>test_reliability_voxcpm2.py</code>)
-      </td>
-      <td>
-        <a href="#chapter-4-l5-level-testing-stability-and-reliability-testing">Chapter 4</a><br>
-        L5: Purpose, Test Content, Directory Location, Example
-      </td>
-      <td>Weekly / Days before Release</td>
-      <td>GPU</td>
-    </tr>
-  </tbody>
-</table>
-
-
----
-<details>
-<summary> The folder structure for tests file based on the 5 levels design</summary>
 Legend: `✅` = test exists, `⬜` = suggested to add.
 ```
 vllm_omni/                                    tests/
@@ -304,134 +136,219 @@ vllm_omni/                                    tests/
 # vllm_omni/deploy/ instead — see docs/configuration/stage_configs.md.
 ```
 
+## Markers for Tests
 
-</details>
+By adding markers before test functions, tests can later be executed uniformly by simply declaring the corresponding marker type.
 
+### Current Markers
+Defined in `pyproject.toml`:
 
-## Common Specifications
+| Marker             | Description                                               |
+| ------------------ | --------------------------------------------------------- |
+| `core_model`       | L1&L2 tests (run in each PR)                              |
+| `advanced_model`   | L3 tests (run on each merge to main)                 |
+| `full_model`       | L4 tests (run nightly) |
+| `diffusion`        | Diffusion model tests                                     |
+| `omni`             | Omni multimodal model tests |
+| `tts`              | TTS model tests     |
+| `cache`            | Cache backend tests                                       |
+| `parallel`         | Parallelism/distributed tests                             |
+| `cpu`              | Tests that run on CPU                                     |
+| `gpu`              | Tests that run on GPU *                                   |
+| `cuda`             | Tests that run on CUDA *                                  |
+| `rocm`             | Tests that run on AMD/ROCm *                              |
+| `xpu`              | Tests that run on Intel XPU *                             |
+| `npu`              | Tests that run on NPU/Ascend *                            |
+| `H100`             | Tests that require H100 GPU  *                            |
+| `L4`               | Tests that require L4 GPU *                               |
+| `MI325`            | Tests that require MI325 GPU (AMD/ROCm) *                 |
+| `A2`               | Tests that require A2 NPU *                               |
+| `A3`               | Tests that require A3 NPU *                               |
+| `distributed_cuda` | Tests that require multi cards on CUDA platform *         |
+| `distributed_rocm` | Tests that require multi cards on ROCm platform  *        |
+| `distributed_npu`  | Tests that require multi cards on NPU platform  *         |
+| `skipif_cuda`      | Skip if the num of CUDA cards is less than the required * |
+| `skipif_rocm`      | Skip if the num of ROCm cards is less than the required * |
+| `skipif_npu`       | Skip if the num of NPU cards is less than the required *  |
+| `slow`             | Slow tests (may skip in quick CI)                         |
+| `benchmark`        | Benchmark tests (decorator on runner test functions; perf JSON uses `full_model` + type marker instead) |
+| `local_model`      | Tests requiring local / non-HF-hub model weights          |
 
-Before entering specific testing levels, the project establishes two common specifications aimed at standardizing the development process and quickly locating issues.
+\* Means those markers are auto-added by `@hardware_test` (parametrization decorator) or `hardware_marks` (only returning the list of marks for flexibility).
 
-1.  ***PR Checklist ([Tests Style](../ci/tests_style.md))***: This template defines the self-check items that must be completed before submitting a code review (Pull Request). It ensures that each code change meets basic requirements such as code style, dependency updates, and documentation synchronization before entering the automated testing pipeline, serving as the first manual line of defense for quality assurance.
-2.  ***CI Failure Explanation ([CI Failures](../ci/failures.md))***: This document archives and explains common failure patterns in the Continuous Integration (CI) pipeline, error log interpretation, and preliminary troubleshooting steps. It helps developers and testers quickly diagnose the causes of automated test failures, improving problem-solving efficiency.
+#### Example usage for markers
 
-## Notes
+```python
+from tests.helpers.mark import hardware_test
 
-### Diff-aware Buildkite uploads (`source_file_dependencies`)
-
-L2 (`.buildkite/test-ready.yml`) and L3 (`.buildkite/test-merge.yml`) pipelines can **skip unrelated GPU jobs at upload time** based on the PR diff. This is implemented by `.buildkite/scripts/upload_pipeline.py`, which filters steps before calling `buildkite-agent pipeline upload`.
-
-#### What `source_file_dependencies` is
-
-- A **uploader-only** YAML key on a Buildkite step or group. **Buildkite does not understand it**; `upload_pipeline.py` always **removes** it from the YAML that is uploaded.
-- A list of path **prefixes** (directories or individual files). If **any** changed file in the diff equals a prefix or starts with `prefix/`, the step is kept; otherwise the step (or entire group) is **omitted** from the uploaded pipeline.
-
-#### When filtering runs
-
-| Build context | Changed files used for matching |
-| --- | --- |
-| Pull request | `git diff --name-only origin/<base>...<BUILDKITE_COMMIT>` |
-| `main` branch push | `git diff --name-only <commit>^..<commit>` |
-| Other (e.g. local dry-run, non-PR branch) | Diff cannot be resolved → **no filtering**; all steps are uploaded and `source_file_dependencies` is still stripped |
-
-Docs-only PRs are handled earlier in bootstrap (`pipeline.yml`) via skip-ci logic; `source_file_dependencies` applies only to the **uploaded** L2/L3 test pipelines.
-
-#### Where it is configured
-
-| Level | Pipeline file | Upload entry |
-| --- | --- | --- |
-| L2 | `.buildkite/test-ready.yml` | `upload_pipeline.py --upload .buildkite/test-ready.yml` (from `pipeline.yml`) |
-| L3 | `.buildkite/test-merge.yml` | `upload_pipeline.py --upload .buildkite/test-merge.yml` (from `pipeline.yml`) |
-
-Steps **without** `source_file_dependencies` are always uploaded (subject to the usual label conditions: `ready` for L2, `merge-test` for L3).
-
-#### Current skip policy (L2 / L3)
-
-To balance CI cost and coverage:
-
-- **Always run** (no `source_file_dependencies`): baseline groups outside E2E Test—e.g. Simple Test, Diffusion unit tests, Engine/Model Executor, Distributed, Custom Pipeline, Entrypoints (L2), LoRA / Entrypoints (L3).
-- **Diff-gated** (`source_file_dependencies` set): **every leaf job under the E2E Test group** in `test-ready.yml` and `test-merge.yml`, regardless of queue (`mithril-h100-pool`, `gpu_1_queue`, or `gpu_4_queue`). Each step lists the smallest set of prefixes that should trigger it—typically:
-  - pytest file(s) exercised by the job (online and/or offline);
-  - model code under `vllm_omni/model_executor/models/` or `vllm_omni/diffusion/models/`;
-  - related `vllm_omni/model_executor/stage_input_processors/` and `vllm_omni/deploy/*.yaml` when applicable.
-
-Adding a new E2E step: add `source_file_dependencies` on the leaf job with those prefixes. Prefer **per-step** deps rather than a broad group-level list unless every child shares the same paths.
-
-#### YAML examples
-
-H100 E2E (kubernetes / `mithril-h100-pool`):
-
-```yaml
-      - label: "Diffusion · Qwen Image Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen_image.py
-          - vllm_omni/diffusion/models/qwen_image/
-        commands:
-          - pytest -s -v tests/e2e/online_serving/test_qwen_image.py -m 'core_model' ...
-        agents:
-          queue: "mithril-h100-pool"
+@pytest.mark.core_model
+@pytest.mark.omni
+@hardware_test(
+   res={"cuda": "L4", "rocm": "MI325", "npu": "A2"},
+   num_cards=2,
+)
+@pytest.mark.parametrize("omni_server", test_params, indirect=True)
+def test_video_to_audio()
+    ...
 ```
 
-Docker E2E (`gpu_1_queue` / `gpu_4_queue`)—same key, same upload-time filtering:
+#### Decorator: `@hardware_test`
 
-```yaml
-      - label: "TTS · Qwen3-TTS CustomVoice Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen3_tts_customvoice.py
-          - vllm_omni/model_executor/models/qwen3_tts/
-          - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
-          - vllm_omni/deploy/qwen3_tts.yaml
-        commands:
-          - pytest -s -v tests/e2e/online_serving/test_qwen3_tts_customvoice.py ...
-        agents:
-          queue: "gpu_4_queue"
+This decorator is intended to make hardware-aware, cross-platform test authoring easier and more robust for CI/CD environments. The `hardware_test` decorator in `vllm-omni/tests/helpers/mark.py` performs the following actions:
+
+1. **Applies platform and resource markers**  
+   Adds the appropriate pytest markers for each specified hardware platform (e.g., `cuda`, `rocm`, `xpu`, `npu`) and resource type (e.g., `L4`, `H100`, `MI325`, `B60`, `A2`, `A3`).
+   ```
+   @pytest.mark.cuda
+   @pytest.mark.L4
+   ```
+2. **Handles multi-card (distributed) scenarios**  
+   For tests requiring multiple cards, it automatically adds distributed markers such as `distributed_cuda`, `distributed_rocm`, or `distributed_npu`.
+   ```
+   @pytest.mark.distributed_cuda(num_cards=num_cards)
+   ```
+3. **Supports flexible card requirements**  
+   Accepts `num_cards` as either a single integer for all platforms or as a dictionary with per-platform values. If not specified, defaults to 1 card per platform.
+
+4. **Integrates resource validation**  
+   On CUDA, adds a skip marker (`skipif_cuda`) if the system does not have the required number of devices.
+   Support for `skipif_rocm` and `skipif_npu` will be implemented later.
+
+
+5. **Works with pytest filtering**  
+   Allows tests to be filtered and selected at runtime using standard pytest marker expressions (e.g., `-m "distributed_cuda and L4"`).
+
+##### Example usage for decorator
+- Single call for multiple platforms:
+    ```python
+    @hardware_test(
+        res={"cuda": "L4", "rocm": "MI325", "xpu": "B60", "npu": "A2"},
+        num_cards={"cuda": 2, "rocm": 2, "xpu": 2, "npu": 2},
+    )
+    ```
+    or
+    ```python
+    @hardware_test(
+        res={"cuda": "L4", "rocm": "MI325", "xpu": "B60", "npu": "A2"},
+        num_cards=2,
+    )
+    ```
+- `res` must be a dict; supported resources: CUDA (L4/H100), ROCm (MI325), XPU (B60), MUSA (S5000), NPU (A2/A3)
+- `num_cards` can be int (all platforms) or dict (per platform); defaults to 1 when missing
+- Distributed markers (`distributed_cuda`, `distributed_rocm`, `distributed_npu`) are auto-added for multi-card cases
+- Filtering examples:
+    - CUDA only: `pytest -m "distributed_cuda and L4"`
+    - ROCm only: `pytest -m "distributed_rocm and MI325"`
+    - NPU only: `pytest -m "distributed_npu"`
+
+#### Function: `hardware_marks`
+
+`hardware_marks` returns a list of pytest mark objects with the same signature as `@hardware_test`. Use it when you need more flexibility, such as attaching hardware marks to individual `pytest.param` entries rather than an entire test function.
+
+```python
+from tests.helpers.mark import hardware_marks
+
+MULTI_CARD_MARKS = hardware_marks(
+    res={"cuda": "H100", "rocm": "MI325", "npu": "A2"}, num_cards=2
+)
+
+@pytest.mark.parametrize("omni_server", [
+    pytest.param(OmniServerParams(...), id="case_001", marks=MULTI_CARD_MARKS),
+], indirect=True)
+def test_feature(omni_server):
+    ...
 ```
 
-A **group** may also define `source_file_dependencies`; nested steps inherit filtering as a unit—the whole group is dropped if no prefix matches.
+#### JSON `mark` field (L4 perf configs)
 
-#### Local dry-run
+Perf JSON under `tests/dfx/perf/tests/` can attach pytest marks per **case** (`test_name` block). Parsed by `tests.dfx.conftest.resolve_pytest_marks` and applied to each parametrized run in `run_benchmark.py` / `run_diffusion_benchmark.py`.
 
-```bash
-# Render filtered YAML to stdout (no upload)
-python3 .buildkite/scripts/upload_pipeline.py .buildkite/test-ready.yml
+When `mark` is present, it must be an **array** with exactly one ``hardware_marks`` object (same semantics as `hardware_marks()` above), followed by registered pytest marker name strings such as `full_model`, `omni`, `tts`, or `diffusion`.
 
-# Confirm uploader-only keys are stripped
-python3 .buildkite/scripts/upload_pipeline.py .buildkite/test-merge.yml | grep source_file_dependencies
-# (no output expected)
+```json
+{
+  "test_name": "test_cosmos3_t2i_official_demo_2gpu",
+  "mark": [
+    {"hardware_marks": {"res": {"cuda": "H100"}, "num_cards": 2}},
+    "full_model",
+    "diffusion"
+  ],
+  "server_type": "vllm-omni",
+  "server_params": { "...": "..." },
+  "benchmark_params": [{ "name": "1024x1024_steps4", "...": "..." }]
+}
 ```
 
-On a PR build, Buildkite logs from `upload_pipeline.py` include lines such as `skip '…' (no changes under …)` for omitted steps.
+- Local bulk load: `pytest -sv tests/dfx/perf/scripts/run_diffusion_benchmark.py -m "full_model and diffusion and H100"`
+- Nightly CI perf steps: `--test-config-file tests/dfx/perf/tests/test_<model>_vllm_omni.json` (file selects cases; no `-m`)
+- Result filenames use **runtime** GPU detection (`get_runtime_resource_label`); `H100` is omitted on the default CI pool
 
-#### Related
+See [L4 performance test examples](./test_examples/l4_performance_tests.inc.md) for the full schema.
 
-- Implementation: `.buildkite/scripts/upload_pipeline.py`
-- L2/L3 diff skip does **not** replace label-based triggers (`ready`, `merge-test`); it only reduces which steps appear **after** the pipeline is already scheduled.
+### Add Support for a New Platform
 
-### Test helper environment variables
+If you want to add support for a new platform (e.g., "tpu" for a new accelerator), follow these steps:
 
-Some shared helpers under `tests/helpers/` honor optional environment variables for local debugging. These are **not** set in CI by default.
+1. **Extend the marker list in your pytest config** so that platform/resource markers are defined:
+   ```toml
+   # In pyproject.toml or pytest.ini
+   [tool.pytest.ini_options]
+   markers = [
+       # ... existing markers ...
+       "tpu: Tests that require TPU device",
+       "TPU_V3: Tests that require TPU v3 hardware",
+       "distributed_tpu: Tests that require multiple TPU devices",
+   ]
+   ```
+2. **Implement a marker construction function for your platform** in `vllm-omni/tests/helpers/mark.py`:
+   ```python
+   # In vllm-omni/tests/helpers/mark.py
 
-| Variable | Accepted values | Description |
-| -------- | --------------- | ----------- |
-| `VLLM_OMNI_KEEP_REQUEST_MEDIA` | `1`, `true`, `yes` (case-insensitive) | When enabled, temporary WAV files created by `tests.helpers.media.convert_audio_bytes_to_text` are **not** deleted when the pytest process exits. By default, each call writes a unique file under the system temp directory via `tempfile.mkstemp` and registers `atexit` cleanup. Use this when debugging audio output validation (Whisper transcription, keyword checks, text–audio similarity). The saved path is logged as `audio data is saved: <path>`. |
+   def tpu_marks(*, res: str, num_cards: int):
+       test_platform = pytest.mark.tpu
+       if res == "TPU_V3":
+           test_resource = pytest.mark.TPU_V3
+       else:
+           raise ValueError(
+               f"Invalid TPU resource type: {res}. Supported: TPU_V3")
 
-Example (Linux / macOS):
+       if num_cards == 1:
+           return [test_platform, test_resource]
+       else:
+           test_distributed = pytest.mark.distributed_tpu(num_cards=num_cards)
+           # Optionally: add skipif_tpu when implemented
+           return [test_platform, test_resource, test_distributed]
+   ```
+3. **Update `hardware_marks` to recognize your new platform**:
+    In the relevant place (see the `hardware_marks` implementation), add:
+    ```python
+    if platform == "tpu":
+        marks = tpu_marks(res=resource, num_cards=cards)
+    ```
+    (`hardware_test` calls `hardware_marks` internally, so both will pick up the change.)
+4. **(Recommended) Add a test using your new markers**:
+   ```python
+   @hardware_test(
+       res={"tpu": "TPU_V3"},
+       num_cards=2,
+   )
+   def test_my_tpu_feature():
+       ...
+   ```
 
-```bash
-export VLLM_OMNI_KEEP_REQUEST_MEDIA=1
-pytest -s -v tests/e2e/online_serving/test_qwen3_omni.py -k test_mix_to_text_audio
-```
+**Summary**:  
+- Add pytest markers for your new platform/resources  
+- Implement a marker function (`xxx_marks`)  
+- Plug into `hardware_marks`  
+- You're done: tests using `@hardware_test` or `hardware_marks` with your platform now automatically get the correct markers, distribution, and isolation!
 
-Example (Windows PowerShell):
+See code in `vllm-omni/tests/helpers/mark.py` for existing examples (`cuda_marks`, `rocm_marks`, `npu_marks`).
 
-```powershell
-$env:VLLM_OMNI_KEEP_REQUEST_MEDIA = "1"
-pytest -s -v tests/e2e/online_serving/test_qwen3_omni.py -k test_mix_to_text_audio
-```
+## Test case style
 
-## Chapter 1: L1 & L2 Level Testing - Unit Testing and Basic End-to-End Verification
+### L1 & L2 Level Testing - Unit Testing and Basic End-to-End Verification
 
-### 1.1 Testing Purpose
+#### 1.1 Testing Purpose
 
 L1 and L2 level testing form the foundation of the quality assurance system. L1 level testing focuses on verifying the internal logic correctness of code units (e.g., functions, classes), ensuring each independent component behaves as designed.
 
@@ -439,7 +356,7 @@ L2 level testing builds upon L1 by introducing GPU resources and verifying that
 
 
 
-### 1.2 Testing Content and Scope
+#### 1.2 Testing Content and Scope
 
 -   ***L1 (Unit & Logic Testing)***:
 -   -   ***Scope***: Tests internal functions and methods of core components such as `entrypoints`, `models`.
@@ -454,7 +371,7 @@ L2 level testing builds upon L1 by introducing GPU resources and verifying that
         -   **Omni multimodal** (`assert_omni_response`): L2 asserts successful completion; keyword, transcript, and cross-modal similarity checks run only at L3+.
     -   ***Characteristic***: Requires ***GPU*** resources to perform model computations.
 
-### 1.3 Test Directory and Execution Files
+#### 1.3 Test Directory and Execution Files
 
 A clear directory structure is key to managing test cases efficiently.
 
@@ -464,10 +381,10 @@ A clear directory structure is key to managing test cases efficiently.
 -   -   Online Serving: `/tests/e2e/online_serving/test_{model_name}.py`
     -   Offline Inference: `/tests/e2e/offline_inference/test_{model_name}.py`
 
-### 1.4 Execution Method and Example
+#### 1.4 Execution Method and Example
 
 -   ***Trigger Timing***: **`PR with ready label`**. That is, when a developer adds a "ready for review" or similar label to a PR on platforms like GitHub, L1 and L2 tests are automatically triggered.
--   ***Diff-aware step skipping***: On L2, **E2E Test** jobs may be omitted at pipeline upload when the PR diff does not touch their [`source_file_dependencies`](#diff-aware-buildkite-uploads-source_file_dependencies) prefixes (H100 and docker queues alike); non-E2E groups still always upload. See [Diff-aware Buildkite uploads](#diff-aware-buildkite-uploads-source_file_dependencies).
+-   ***Diff-aware step skipping***: On L2, **E2E Test** jobs may be omitted at pipeline upload when the PR diff does not touch their [`source_file_dependencies`](ci_settings.md#step-filtering) prefixes; non-E2E groups still always upload. See [CI Settings — Diff-aware CI](ci_settings.md#diff-aware-ci).
 -   ***Execution Environment***: L1 uses ***CPU*** environment; L2 requires ***GPU*** environment.
 -   ***Script Example***:
 
@@ -535,7 +452,7 @@ def test_cap_and_align_mel_length_no_mismatch(repeats, code_len, max_mel_frames)
 
 <details>
 <summary> L2 Test Examples</summary>
-You can refer to Test Examples in Chapter 2 to see example test cases that incorporate both L2 and L3 testing logic.
+You can refer to Test Examples in the L3 section to see example test cases that incorporate both L2 and L3 testing logic.
 </details>
 
 -   -   ***Run Command***:
@@ -543,9 +460,9 @@ You can refer to Test Examples in Chapter 2 to see example test cases that incor
     `pytest -s -v /tests/e2e/online_serving/test_{model_name}.py`
     `pytest -s -v -m 'core_model and cpu' --run-level=core_model`
 
-## Chapter 2: L3 Level Testing - Core Integration, Performance, and Accuracy Verification
+### L3 Level Testing - Core Integration, Performance, and Accuracy Verification
 
-### 2.1 Testing Purpose
+#### 2.1 Testing Purpose
 
 L3 level testing executes after code is merged into the main branch. Its core purpose is to verify the integration effect, key performance indicators, and output accuracy of ***real models*** in ***multiple deployment scenarios***
 
@@ -553,30 +470,30 @@ L3 level testing executes after code is merged into the main branch. Its core pu
 
 
 
-### 2.2 Testing Content and Scope
+#### 2.2 Testing Content and Scope
 
 -   ***Deployment Scenarios***: Covers richer `online` and `offline` deployment configurations, which may include different hardware configurations, batch sizes, concurrency levels, etc.
 -   ***Core Verification***:
 -   1.  ***Inference Functionality***: Ensures real models can perform forward computation normally and return results.
     2.  ***Accuracy Compliance***: Verifies that the model's evaluation metrics (e.g., accuracy) meet the expected baseline, preventing code changes from introducing accuracy issues.
     3.  ***Important Performance***: Verifies whether performance (e.g., P99 latency, throughput) in core scenarios meets preset thresholds.
--   ***L3 response validation (`advanced_model`)***: At `--run-level advanced_model`, `tests/helpers/assertions.py` adds semantic checks on top of L2 payload gates (see Chapter 1):
+-   ***L3 response validation (`advanced_model`)***: At `--run-level advanced_model`, `tests/helpers/assertions.py` adds semantic checks on top of L2 payload gates (see L1 & L2 above):
     -   **Speech / TTS** (`assert_audio_speech_response`): Whisper transcript of returned audio vs `request_config["input"]` (cosine similarity &gt; 0.9 when input text is set); optional `min_audio_bytes` floor; PCM harmonic-to-noise ratio when `response_format` is `pcm`; **preset voice gender** via `_assert_preset_voice_gender_from_audio` when `voice` matches a known preset in `_PRESET_VOICE_GENDER_MAP` (pitch/F0-based classifier; skipped for unknown voices or `pcm` output).
     -   **Omni multimodal** (`assert_omni_response`): non-empty text/audio outputs; `key_words` in transcript or text; text–audio similarity / containment; preset **`speaker`** gender check (same helper as TTS).
     -   **Diffusion** (`assert_*_diffusion_response`): image/video dimension and frame-count parity with request parameters where configured.
     -   **Accuracy suites**: pixel/video similarity and metric baselines under `/tests/e2e/accuracy/` (separate from inline helper assertions).
 
-### 2.3 Test Directory and Execution Files
+#### 2.3 Test Directory and Execution Files
 
 -   ***Functional Testing***:
 -   -   Online Serving: `/tests/e2e/online_serving/test_{model_name}_expansion.py`
     -   Offline Inference: `/tests/e2e/offline_inference/test_{model_name}_expansion.py`
     -   (Note: `_expansion.py` likely means it contains more comprehensive scenario cases compared to L2 tests).
 
-### 2.4 Execution Method and Example
+#### 2.4 Execution Method and Example
 
 -   ***Trigger Timing***: **`PR Merged`**. Automatically triggered after code review is approved and merged into the main branch (typically via `merge-test` label on the PR before merge).
--   ***Diff-aware step skipping***: Same [`source_file_dependencies`](#diff-aware-buildkite-uploads-source_file_dependencies) mechanism as L2—**all E2E Test** leaf steps are diff-gated; other L3 groups always upload. See [Diff-aware Buildkite uploads](#diff-aware-buildkite-uploads-source_file_dependencies).
+-   ***Diff-aware step skipping***: Same [`source_file_dependencies`](ci_settings.md#step-filtering) mechanism as L2—**all E2E Test** leaf steps are diff-gated; other L3 groups always upload. See [CI Settings — Diff-aware CI](ci_settings.md#diff-aware-ci).
 -   ***Execution Environment***: ***GPU*** servers.
 -   ***Script Example***:
 
@@ -692,27 +609,27 @@ L3 level testing executes after code is merged into the main branch. Its core pu
 
     **Implicit Validation**: `send_omni_request`, `send_audio_speech_request`, and `send_diffusion_request` call `assert_*_response` helpers with the session `--run-level`. At `core_model`, checks are limited to success plus cheap payload sanity (non-empty audio/media, format/content-type, optional `min_audio_bytes`). At `advanced_model` and `full_model`, deep validation adds Whisper transcripts, keyword/similarity, diffusion dimension checks, PCM HNR, etc.
 
-    **Audio output debugging**: Deep validation may transcribe returned audio via `convert_audio_bytes_to_text` (Whisper). If an audio keyword or text–audio similarity assertion fails, set `VLLM_OMNI_KEEP_REQUEST_MEDIA=1` before running pytest to keep the intermediate WAV files for inspection (see [Test helper environment variables](#test-helper-environment-variables)).
+    **Audio output debugging**: Deep validation may transcribe returned audio via `convert_audio_bytes_to_text` (Whisper). If an audio keyword or text–audio similarity assertion fails, set `VLLM_OMNI_KEEP_REQUEST_MEDIA=1` before running pytest to keep the intermediate WAV files for inspection (see [Test helper environment variables](test_system_overview.md#test-helper-environment-variables)).
 
 -   ***Run Command (L3 merge)***: `pytest -s -v /tests/e2e/online_serving/test_{model_name}.py -m advanced_model --run-level=advanced_model`
 
 -   ***Run Command (L4 nightly expansion)***: `pytest -s -v /tests/e2e/online_serving/test_{model_name}_expansion.py -m full_model --run-level=full_model`
 
-## Chapter 3: L4 Level Testing - Full Functionality, Performance, and Documentation Testing
+### L4 Level Testing - Full Functionality, Performance, and Documentation Testing
 
-### 3.1 Testing Purpose
+#### 3.1 Testing Purpose
 
 L4 level testing is a comprehensive quality audit before a version release. It expands upon L3, executing ***full*** functional scenarios, conducting systematic ***performance stress tests***, and simultaneously verifying the correctness of accompanying ***example documentation***. Its purpose is to perform deep validation of the system during off-peak nighttime hours, providing quality trend reports for daytime development and data support for release decisions.
 
 
 
-### 3.2 Testing Content and Scope
+#### 3.2 Testing Content and Scope
 
 -   ***Full Functionality Testing***: Executes all test cases defined in `test_{model_name}_expansion.py`, covering all implemented features, positive flows, boundary conditions, and exception handling.
 -   ***Performance Testing***: Uses `tests/dfx/perf/tests/test_qwen3_omni_*.json` (Omni), `test_tts.json` / `test_voxcpm2.json` / `test_higgs_audio_v3.json` (TTS), and diffusion configs `tests/dfx/perf/tests/test_*_vllm_omni.json` (passed to `run_benchmark.py` or `run_diffusion_benchmark.py` via `--test-config-file` in nightly **Perf Test** steps) to drive throughput, latency, and memory benchmarks. Each JSON **case** may declare an optional top-level **`mark`** array: exactly one ``hardware_marks`` object plus pytest marker name strings (`full_model`, `omni` / `tts` / `diffusion`, …). Runners attach those marks to each parametrized `(server, benchmark index)` pair so **local** bulk runs can filter with `-m` (for example `-m "full_model and H100 and diffusion"`). Nightly CI perf jobs select workloads by **`--test-config-file`**, not `-m`. Details are in the Performance Tests example below.
 -   ***Documentation Testing***: Verifies whether the example code provided to users is runnable and its results match the description.
 
-### 3.3 Test Directory and Execution Files
+#### 3.3 Test Directory and Execution Files
 
 -   ***Functional Testing***: Same directories as L3.
 -   ***Performance Test Configuration***: `tests/dfx/perf/tests/test_qwen3_omni_*.json`, `test_tts.json`, `test_voxcpm2.json`, `test_higgs_audio_v3.json`, and diffusion configs `tests/dfx/perf/tests/test_*_vllm_omni.json` (e.g. `test_qwen_image_vllm_omni.json`, `test_cosmos3_vllm_omni.json`). Optional per-case `mark` for local `-m` filtering is documented in Section 3.4 Performance Tests.
@@ -720,7 +637,7 @@ L4 level testing is a comprehensive quality audit before a version release. It e
     -   `tests/examples/online_serving/test_{model_name}.py`
     -   `tests/examples/offline_inference/test_{model_name}.py`
 
-### 3.4 Execution Method and Example
+#### 3.4 Execution Method and Example
 
 -   ***Trigger Timing***: **`Nightly`**, automatically executed every night.
 -   ***Execution Environment***: ***GPU*** server clusters to meet the resource demands of performance testing.
@@ -740,20 +657,20 @@ L4 level testing is a comprehensive quality audit before a version release. It e
 
 -   ***Run Command***: (Specific commands would depend on the performance testing tool and configuration defined in `nightly.json`).
 
-## Chapter 4: L5 Level Testing - Stability and Reliability Testing
+### L5 Level Testing - Stability and Reliability Testing
 
-### 4.1 Testing Purpose
+#### 4.1 Testing Purpose
 
 L5 level testing focuses on the performance of model services under ***long-running*** and ***abnormal fault*** scenarios. It aims to uncover deep-seated issues that only manifest under sustained pressure or extreme conditions, such as memory leaks, resource contention, gradual performance degradation, and lack of fault tolerance mechanisms. This is the final, yet crucial, line of defense for ensuring service high availability and production environment robustness.
 
 
 
-### 4.2 Testing Content and Scope
+#### 4.2 Testing Content and Scope
 
 -   ***Long-term Stability (Stability) Testing***: Uses JSON under `tests/dfx/stability/tests/` (for example `test_qwen3_omni.json` and `test_wan22.json`) to run the service under moderate load for an extended period (e.g., over 12 hours), monitoring whether metrics like memory/VRAM usage, response time, and throughput degrade over time, and whether the service process remains stable.
 -   ***Reliability Testing***: Uses pytest suites under `tests/dfx/reliability/` to inject controlled faults against a **live** `vllm_omni serve` instance (same **`omni_server` / `omni_server_function`** fixture style as E2E). Current suites emphasize **GPU memory pressure** (CUDA sidecar “memory hog”), **worker / runtime process kill** (`SIGKILL` on `VLLM::Worker` for Qwen3-Omni, `multiprocessing.spawn` for Wan2.2 video workers, or `vLLM-Omni::` for HunyuanImage DiT workers), **large multimodal chat**, **`/v1/videos`**, **`/v1/images/generations`**, or **`/v1/audio/speech`** jobs under OOM, **`/health` → 503** and **fast-fail / non-hanging concurrent** requests after kill, and **OpenAI-style 5xx error contracts** (e.g. text vs text+audio under OOM). **Post-fault recovery** checks exist where enabled (some cases may be `skip` while issues are tracked). See the Reliability `<details>` block in Section 4.4 for file-level responsibilities and CI markers (`slow`, `hardware_test`, POSIX-only kill).
 
-### 4.3 Test Directory and Execution Files
+#### 4.3 Test Directory and Execution Files
 
 -   ***Stability Test Configuration***: `tests/dfx/stability/tests/test_qwen3_omni.json`, `tests/dfx/stability/tests/test_wan22.json` (one JSON per model / runner family)
 -   ***Reliability Test Suite*** (`tests/dfx/reliability/`):
@@ -765,7 +682,7 @@ L5 level testing focuses on the performance of model services under ***long-runn
     -   `conftest.py` — `fault_injector` and `omni_server_after_fault` / `omni_server_after_fault_function` fixtures to run a callable **after** the server is ready.
     -   `README.md` — Short local run commands for this directory.
 
-### 4.4 Execution Method and Example
+#### 4.4 Execution Method and Example
 
 -   ***Trigger Timing***: **`Weekly`** (weekly) or **`Days before Release`** (several days before a major release). Due to long execution times, the frequency is lower.
 -   ***Execution Environment***: ***GPU*** servers, requiring a stable and exclusive testing environment.
@@ -799,7 +716,7 @@ When you want to add L5-level stability test cases, add or extend the appropriat
 }
 ```
 
-#### Parameter Explanation
+##### Parameter Explanation
 
 ***Overview***
 
@@ -809,16 +726,16 @@ When you want to add L5-level stability test cases, add or extend the appropriat
 | server_params    | Yes      | Server-side configuration parameters (model, stage configuration, etc.)     |
 | benchmark_params | Yes      | Stability benchmark running parameters (supports multiple configurations)   |
 
-#### server_params Configuration
+##### server_params Configuration
 
-##### Basic Parameters
+###### Basic Parameters
 
 | Parameter         | Required | Example                            | Description                         |
 | ----------------- | -------- | ---------------------------------- | ----------------------------------- |
 | model             | Yes      | "Qwen/Qwen3-Omni-30B-A3B-Instruct" | Model name or path                  |
 | stage_config_name | Yes      | "qwen3_omni.yaml"                  | Stage configuration file name       |
 
-##### Dynamic Configuration (update/delete)
+###### Dynamic Configuration (update/delete)
 
 Supports incremental modifications based on the basic configuration:
 
@@ -828,9 +745,9 @@ Supports incremental modifications based on the basic configuration:
 | delete    | Delete specified configuration items |
 
 ***Example***:
-You can refer to Test Examples in Chapter 3.4
+You can refer to Test Examples in L4 §3.4
 
-#### benchmark_params Configuration
+##### benchmark_params Configuration
 
 For stability testing, the key parameters are:
 
@@ -840,18 +757,18 @@ For stability testing, the key parameters are:
     -   `max_concurrency`: Maximum number of concurrent requests. When this is used, `request_rate` is set to `inf` internally.
 -   **num_prompts_per_batch**: Number of prompts sent in each batch. Multiple batches will be executed sequentially within `duration_sec`.
 
-All other optional parameters follow the same rules as the in Chapter 3.4.
+All other optional parameters follow the same rules as in L4 §3.4.
 
 </details>
 
 <details>
 <summary> Reliability test suite (<code>tests/dfx/reliability</code>)</summary>
 
-#### Purpose and relationship to stability
+##### Purpose and relationship to stability
 
 Reliability tests are **short fault-injection** integration runs (L5 **(b)** in `tests/dfx/reliability/README.md`). They complement **stability** JSON-driven long runs: instead of hours of steady traffic, they **perturb** the server (GPU OOM sidecar, fatal signals on selected processes) and check **failure mode** and **latency bounds** (e.g. chat or `/v1/videos` must not hang under concurrent fault-time load).
 
-#### Directory layout
+##### Directory layout
 
 | Path | Responsibility |
 | ---- | -------------- |
@@ -863,15 +780,15 @@ Reliability tests are **short fault-injection** integration runs (L5 **(b)** in
 | `test_reliability_voxcpm2.py` | VoxCPM2: `/v1/audio/speech` under OOM (error contract), **SIGKILL** on `VLLM::` workers and serve/tree targets, health / fast-fail / concurrent speech requests; some OOM cases may be skipped while tracked in issues. |
 | `README.md` | Minimal run / collect examples. |
 
-#### Parametrization and markers
+##### Parametrization and markers
 
 - Each test module defines a **`RELIABILITY_SCENARIOS`** list (`test_name`, `server_params`: model, `stage_config_name` or diffusion `server_args`, etc.). **`create_reliability_omni_server_params()`** in `tests/dfx/conftest.py` resolves stage paths (including XPU substitutions where applicable) and builds **`OmniServerParams`** lists consumed by **`@pytest.mark.parametrize(..., indirect=True)`** on `omni_server` or `omni_server_function`.
 - Cases are tagged **`@pytest.mark.slow`** for weekly / selective CI. GPU-heavy suites use **`@hardware_test(res={"cuda": "H100"}, num_cards=...)`** or **`@hardware_test(res={"cuda": "L4"}, num_cards=1)`** (Qwen3-Omni **2**× H100; Wan2.2 **1**× H100; HunyuanImage DiT **4**× H100; VoxCPM2 **1**× L4).
 - **Process-kill** tests use **`@pytest.mark.skipif(os.name == "nt", ...)`** because injection uses POSIX **`pgrep` / `kill`**.
 
-#### CI trigger
+##### CI trigger
 
-Weekly Buildkite (`.buildkite/test-weekly.yml`) runs one step per model suite (trigger: `WEEKLY=1` or PR label `weekly-test`), for example:
+Weekly Buildkite (`.buildkite/cuda/test-weekly.yml`) runs one step per model suite (trigger: `WEEKLY=1` or PR label `weekly-test`), for example:
 
 | Buildkite step | Test file | CI hardware |
 | -------------- | --------- | ----------- |
@@ -887,7 +804,7 @@ pytest -s -v tests/dfx/reliability/test_reliability_hunyuan_image.py -m "slow"
 pytest -s -v tests/dfx/reliability/test_reliability_voxcpm2.py -m "slow"
 ```
 
-#### Local commands
+##### Local commands
 
 ```bash
 pytest --collect-only tests/dfx/reliability
@@ -897,18 +814,14 @@ pytest -s -v tests/dfx/reliability/test_reliability_hunyuan_image.py -m slow
 pytest -s -v tests/dfx/reliability/test_reliability_voxcpm2.py -m slow
 ```
 
-#### Adding a new model suite
+##### Adding a new model suite
 
 1. Add `test_reliability_<model>.py` under `tests/dfx/reliability/`.
 2. Define **`RELIABILITY_SCENARIOS`** and pass them through **`create_reliability_omni_server_params()`** with the correct deploy or e2e stage-config directory (same pattern as existing files).
 3. Reuse **`helpers`** for OOM / kill / raw HTTP; prefer **`assert_fault_exception()`** and **`resolve_oom_device_spec()`** from `tests/dfx/conftest.py` for consistent device selection vs stage YAML.
-4. Register **`slow`** (and **`hardware_test`** if needed); extend **`.buildkite/test-weekly.yml`** when the suite should run in weekly L5.
+4. Register **`slow`** (and **`hardware_test`** if needed); extend **`.buildkite/cuda/test-weekly.yml`** when the suite should run in weekly L5.
 
 </details>
 
 -   -   ***Stability***: `pytest -s -v tests/dfx/stability/scripts/test_stability_qwen3_omni.py` or `pytest -s -v tests/dfx/stability/scripts/test_stability_wan22.py` (or add `test_stability_<model>.py` alongside a matching JSON config)
-    -   ***Reliability***: `pytest -s -v tests/dfx/reliability/test_reliability_<model>.py -m slow` (current suites: `qwen3_omni`, `wan22`, `hunyuan_image`, `voxcpm2`; add `test_reliability_<suite>.py` and a matching step in `.buildkite/test-weekly.yml` for new models)
-
-## Summary
-
-This multi-level testing system achieves continuous, progressive validation of model service quality by tightly integrating testing activities with the development workflow (commit, review, merge, release). From rapid unit testing to comprehensive end-to-end testing, and further to in-depth performance, stability, and reliability verification, each level has clear objectives, collectively building a robust quality protection net. By following this system, teams can deliver high-quality, highly reliable model services more efficiently.
+    -   ***Reliability***: `pytest -s -v tests/dfx/reliability/test_reliability_<model>.py -m slow` (current suites: `qwen3_omni`, `wan22`, `hunyuan_image`, `voxcpm2`; add `test_reliability_<suite>.py` and a matching step in `.buildkite/cuda/test-weekly.yml` for new models)

docs/contributing/ci/tests_markers.md

@@ -1,207 +0,0 @@
-# Markers for Tests
-
-By adding markers before test functions, tests can later be executed uniformly by simply declaring the corresponding marker type.
-
-## Current Markers
-Defined in `pyproject.toml`:
-
-| Marker             | Description                                               |
-| ------------------ | --------------------------------------------------------- |
-| `core_model`       | L1&L2 tests (run in each PR)                              |
-| `advanced_model`   | L3 tests (run on each merge to main)                 |
-| `full_model`       | L4 tests (run nightly) |
-| `diffusion`        | Diffusion model tests                                     |
-| `omni`             | Omni multimodal model tests |
-| `tts`              | TTS model tests     |
-| `cache`            | Cache backend tests                                       |
-| `parallel`         | Parallelism/distributed tests                             |
-| `cpu`              | Tests that run on CPU                                     |
-| `gpu`              | Tests that run on GPU *                                   |
-| `cuda`             | Tests that run on CUDA *                                  |
-| `rocm`             | Tests that run on AMD/ROCm *                              |
-| `xpu`              | Tests that run on Intel XPU *                             |
-| `npu`              | Tests that run on NPU/Ascend *                            |
-| `H100`             | Tests that require H100 GPU  *                            |
-| `L4`               | Tests that require L4 GPU *                               |
-| `MI325`            | Tests that require MI325 GPU (AMD/ROCm) *                 |
-| `A2`               | Tests that require A2 NPU *                               |
-| `A3`               | Tests that require A3 NPU *                               |
-| `distributed_cuda` | Tests that require multi cards on CUDA platform *         |
-| `distributed_rocm` | Tests that require multi cards on ROCm platform  *        |
-| `distributed_npu`  | Tests that require multi cards on NPU platform  *         |
-| `skipif_cuda`      | Skip if the num of CUDA cards is less than the required * |
-| `skipif_rocm`      | Skip if the num of ROCm cards is less than the required * |
-| `skipif_npu`       | Skip if the num of NPU cards is less than the required *  |
-| `slow`             | Slow tests (may skip in quick CI)                         |
-| `benchmark`        | Benchmark tests (decorator on runner test functions; perf JSON uses `full_model` + type marker instead) |
-| `local_model`      | Tests requiring local / non-HF-hub model weights          |
-
-\* Means those markers are auto-added by `@hardware_test` (parametrization decorator) or `hardware_marks` (only returning the list of marks for flexibility).
-
-### Example usage for markers
-
-```python
-from tests.helpers.mark import hardware_test
-
-@pytest.mark.core_model
-@pytest.mark.omni
-@hardware_test(
-   res={"cuda": "L4", "rocm": "MI325", "npu": "A2"},
-   num_cards=2,
-)
-@pytest.mark.parametrize("omni_server", test_params, indirect=True)
-def test_video_to_audio()
-    ...
-```
-
-### Decorator: `@hardware_test`
-
-This decorator is intended to make hardware-aware, cross-platform test authoring easier and more robust for CI/CD environments. The `hardware_test` decorator in `vllm-omni/tests/helpers/mark.py` performs the following actions:
-
-1. **Applies platform and resource markers**  
-   Adds the appropriate pytest markers for each specified hardware platform (e.g., `cuda`, `rocm`, `xpu`, `npu`) and resource type (e.g., `L4`, `H100`, `MI325`, `B60`, `A2`, `A3`).
-   ```
-   @pytest.mark.cuda
-   @pytest.mark.L4
-   ```
-2. **Handles multi-card (distributed) scenarios**  
-   For tests requiring multiple cards, it automatically adds distributed markers such as `distributed_cuda`, `distributed_rocm`, or `distributed_npu`.
-   ```
-   @pytest.mark.distributed_cuda(num_cards=num_cards)
-   ```
-3. **Supports flexible card requirements**  
-   Accepts `num_cards` as either a single integer for all platforms or as a dictionary with per-platform values. If not specified, defaults to 1 card per platform.
-
-4. **Integrates resource validation**  
-   On CUDA, adds a skip marker (`skipif_cuda`) if the system does not have the required number of devices.
-   Support for `skipif_rocm` and `skipif_npu` will be implemented later.
-
-
-5. **Works with pytest filtering**  
-   Allows tests to be filtered and selected at runtime using standard pytest marker expressions (e.g., `-m "distributed_cuda and L4"`).
-
-#### Example usage for decorator
-- Single call for multiple platforms:
-    ```python
-    @hardware_test(
-        res={"cuda": "L4", "rocm": "MI325", "xpu": "B60", "npu": "A2"},
-        num_cards={"cuda": 2, "rocm": 2, "xpu": 2, "npu": 2},
-    )
-    ```
-    or
-    ```python
-    @hardware_test(
-        res={"cuda": "L4", "rocm": "MI325", "xpu": "B60", "npu": "A2"},
-        num_cards=2,
-    )
-    ```
-- `res` must be a dict; supported resources: CUDA (L4/H100), ROCm (MI325), XPU (B60), MUSA (S5000), NPU (A2/A3)
-- `num_cards` can be int (all platforms) or dict (per platform); defaults to 1 when missing
-- Distributed markers (`distributed_cuda`, `distributed_rocm`, `distributed_npu`) are auto-added for multi-card cases
-- Filtering examples:
-    - CUDA only: `pytest -m "distributed_cuda and L4"`
-    - ROCm only: `pytest -m "distributed_rocm and MI325"`
-    - NPU only: `pytest -m "distributed_npu"`
-
-### Function: `hardware_marks`
-
-`hardware_marks` returns a list of pytest mark objects with the same signature as `@hardware_test`. Use it when you need more flexibility, such as attaching hardware marks to individual `pytest.param` entries rather than an entire test function.
-
-```python
-from tests.helpers.mark import hardware_marks
-
-MULTI_CARD_MARKS = hardware_marks(
-    res={"cuda": "H100", "rocm": "MI325", "npu": "A2"}, num_cards=2
-)
-
-@pytest.mark.parametrize("omni_server", [
-    pytest.param(OmniServerParams(...), id="case_001", marks=MULTI_CARD_MARKS),
-], indirect=True)
-def test_feature(omni_server):
-    ...
-```
-
-### JSON `mark` field (L4 perf configs)
-
-Perf JSON under `tests/dfx/perf/tests/` can attach pytest marks per **case** (`test_name` block). Parsed by `tests.dfx.conftest.resolve_pytest_marks` and applied to each parametrized run in `run_benchmark.py` / `run_diffusion_benchmark.py`.
-
-When `mark` is present, it must be an **array** with exactly one ``hardware_marks`` object (same semantics as `hardware_marks()` above), followed by registered pytest marker name strings such as `full_model`, `omni`, `tts`, or `diffusion`.
-
-```json
-{
-  "test_name": "test_cosmos3_t2i_official_demo_2gpu",
-  "mark": [
-    {"hardware_marks": {"res": {"cuda": "H100"}, "num_cards": 2}},
-    "full_model",
-    "diffusion"
-  ],
-  "server_type": "vllm-omni",
-  "server_params": { "...": "..." },
-  "benchmark_params": [{ "name": "1024x1024_steps4", "...": "..." }]
-}
-```
-
-- Local bulk load: `pytest -sv tests/dfx/perf/scripts/run_diffusion_benchmark.py -m "full_model and diffusion and H100"`
-- Nightly CI perf steps: `--test-config-file tests/dfx/perf/tests/test_<model>_vllm_omni.json` (file selects cases; no `-m`)
-- Result filenames use **runtime** GPU detection (`get_runtime_resource_label`); `H100` is omitted on the default CI pool
-
-See [L4 performance test examples](./test_examples/l4_performance_tests.inc.md) for the full schema.
-
-## Add Support for a New Platform
-
-If you want to add support for a new platform (e.g., "tpu" for a new accelerator), follow these steps:
-
-1. **Extend the marker list in your pytest config** so that platform/resource markers are defined:
-   ```toml
-   # In pyproject.toml or pytest.ini
-   [tool.pytest.ini_options]
-   markers = [
-       # ... existing markers ...
-       "tpu: Tests that require TPU device",
-       "TPU_V3: Tests that require TPU v3 hardware",
-       "distributed_tpu: Tests that require multiple TPU devices",
-   ]
-   ```
-2. **Implement a marker construction function for your platform** in `vllm-omni/tests/helpers/mark.py`:
-   ```python
-   # In vllm-omni/tests/helpers/mark.py
-
-   def tpu_marks(*, res: str, num_cards: int):
-       test_platform = pytest.mark.tpu
-       if res == "TPU_V3":
-           test_resource = pytest.mark.TPU_V3
-       else:
-           raise ValueError(
-               f"Invalid TPU resource type: {res}. Supported: TPU_V3")
-
-       if num_cards == 1:
-           return [test_platform, test_resource]
-       else:
-           test_distributed = pytest.mark.distributed_tpu(num_cards=num_cards)
-           # Optionally: add skipif_tpu when implemented
-           return [test_platform, test_resource, test_distributed]
-   ```
-3. **Update `hardware_marks` to recognize your new platform**:
-    In the relevant place (see the `hardware_marks` implementation), add:
-    ```python
-    if platform == "tpu":
-        marks = tpu_marks(res=resource, num_cards=cards)
-    ```
-    (`hardware_test` calls `hardware_marks` internally, so both will pick up the change.)
-4. **(Recommended) Add a test using your new markers**:
-   ```python
-   @hardware_test(
-       res={"tpu": "TPU_V3"},
-       num_cards=2,
-   )
-   def test_my_tpu_feature():
-       ...
-   ```
-
-**Summary**:  
-- Add pytest markers for your new platform/resources  
-- Implement a marker function (`xxx_marks`)  
-- Plug into `hardware_marks`  
-- You're done: tests using `@hardware_test` or `hardware_marks` with your platform now automatically get the correct markers, distribution, and isolation!
-
-See code in `vllm-omni/tests/helpers/mark.py` for existing examples (`cuda_marks`, `rocm_marks`, `npu_marks`).

docs/contributing/ci/tests_style.md

@@ -1,466 +0,0 @@
-# Test File Structure and Style Guide
-
-To ensure project maintainability and sustainable development, we encourage contributors to submit test code (unit tests, system tests, or end-to-end tests) alongside their code changes. This document outlines the guidelines for organizing and naming test files.
-
-## Checklist before submitting your test files
-
-1. The file is saved in an appropriate place and the file name is clear.
-2. The coding style follows the requirements outlined below.
-3. All test functions have appropriate pytest markers.
-4. For tests that need run in CI, please ensure it labeled as ``@pytest.mark.core_model` the test is configured under the `./buildkite/` folder.
-
-
-## Test Types
-
-For more details about our [Five Levels Tests design](../ci/CI_5levels.md).
-
-### Unit Tests and System Tests
-For unit tests and system tests, we strongly recommend placing test files in the same directory structure as the source code being tested, using the naming convention `test_*.py`.
-
-### End-to-End (E2E) Tests for Models
-End-to-end tests verify the complete functionality of a system or component. For our project, the E2E tests for different omni models are organized into two subdirectories:
-
-- **`tests/e2e/offline_inference/`**: Tests for offline inference modes (e.g., Qwen3Omni offline inference)
-
-- **`tests/e2e/online_serving/`**: Tests for online serving scenarios (e.g., API server tests)
-
-## Test Directory Structure
-
-The ideal directory structure mirrors the source code organization. Legend: `✅` = test exists, `⬜` = suggested to add.
-
-```
-vllm_omni/                                    tests/
-├── config/                             →     ├── config/
-│   ├── model.py                              │   └── test_model.py                    ⬜
-│   └── lora.py                               │   └── test_lora.py                      ⬜
-│
-├── core/                               →     ├── core/
-│   └── sched/                                 │   └── sched/
-│       ├── omni_ar_scheduler.py               │       ├── test_omni_ar_scheduler.py    ⬜
-│       ├── omni_generation_scheduler.py       │       ├── test_omni_generation_scheduler.py  ⬜
-│       └── output.py                          │       └── test_output.py               ✅ currently in entrypoints/test_omni_new_request_data.py (tests output.OmniNewRequestData)
-│
-├── diffusion/                          →     ├── diffusion/
-│   ├── diffusion_engine.py                    │   ├── test_diffusion_engine.py          ⬜
-│   ├── attention/                             │   ├── attention/
-│   │   ├── layer.py                            │   │   ├── test_attention_sp.py         ✅
-│   │   └── backends/                           │   │   └── test_flash_attn.py           ✅
-│   ├── distributed/                           │   ├── distributed/
-│   │   └── ...                                 │   │   ├── test_comm.py                 ✅
-│   │                                            │   │   ├── test_cfg_parallel.py        ✅
-│   │                                            │   │   └── test_sp_plan_hooks.py       ✅
-│   ├── lora/                                   │   ├── lora/
-│   │   └── ...                                 │   │   ├── test_base_linear.py          ✅
-│   │                                            │   │   └── test_lora_manager.py        ✅
-│   ├── models/                                 │   ├── models/
-│   │   ├── qwen_image/                         │   │   ├── qwen_image/                 (e2e coverage)
-│   │   ├── ovis_image/                         │   │   ├── ovis_image/
-│   │   │   └── ...                             │   │   │   └── test_ovis_image.py     ✅
-│   │   ├── z_image/                            │   │   └── z_image/
-│   │   └── ...                                 │   │       └── test_zimage_tp_constraints.py  ✅
-│   └── worker/                                 │   └── worker/
-│       ├── diffusion_worker.py                 │       └── test_diffusion_worker.py   ✅ file at diffusion/test_diffusion_worker.py
-│       └── diffusion_model_runner.py            │
-│
-├── distributed/                         →     ├── distributed/
-│   └── omni_connectors/                         │   └── omni_connectors/
-│       ├── adapter.py                           │       ├── test_adapter_and_flow.py   ✅
-│       ├── kv_transfer_manager.py               │       ├── test_basic_connectors.py   ✅
-│       ├── connectors/                           │       ├── test_kv_flow.py             ✅
-│       └── utils/                               │       └── test_omni_connector_configs.py  ✅
-│
-├── engine/                             →     ├── engine/
-│   ├── input_processor.py                      │   ├── test_input_processor.py         ⬜  (no processor.py in source)
-│   ├── output_processor.py                     │   └── test_output_processor.py         ⬜
-│   └── arg_utils.py                            │   └── test_arg_utils.py               ⬜
-│
-├── entrypoints/                        →     ├── entrypoints/
-│   ├── stage_utils.py                          │   ├── test_stage_utils.py            ✅
-│   ├── cli/                                     │   ├── cli/                           (benchmarks/test_serve_cli.py covers CLI serve)
-│   │   └── ...                                  │   │   └── test_*.py                  ⬜
-│   └── openai/                                  │   └── openai_api/                    # maps to entrypoints/openai/
-│       ├── api_server.py                        │       ├── test_api_server.py         ⬜  (e2e indirect coverage)
-│       ├── serving_chat.py                       │       ├── test_serving_chat_sampling_params.py  ✅
-│       ├── serving_speech.py                     │       ├── test_serving_speech.py     ✅
-│       └── image_api_utils.py                   │       └── test_image_server.py      ✅
-│
-├── inputs/                             →     ├── inputs/
-│   ├── data.py                                 │   ├── test_data.py                   ⬜
-│   ├── parse.py                                │   ├── test_parse.py                 ⬜
-│   └── preprocess.py                            │   └── test_preprocess.py            ✅ currently in entrypoints/test_omni_input_preprocessor.py
-│
-├── model_executor/                     →     ├── model_executor/
-│   ├── layers/                                  │   ├── layers/
-│   │   └── mrope.py                             │   │   └── test_mrope.py              ⬜
-│   ├── model_loader/                            │   ├── model_loader/
-│   │   └── weight_utils.py                      │   │   └── test_weight_utils.py      ⬜
-│   ├── models/                                  │   ├── models/
-│   │   ├── qwen2_5_omni/                         │   │   ├── qwen2_5_omni/
-│   │   │   ├── qwen2_5_omni_thinker.py           │   │   │   ├── test_audio_length.py  ✅
-│   │   │   ├── qwen2_5_omni_talker.py            │   │   │   ├── test_qwen2_5_omni_thinker.py  ⬜
-│   │   │   └── qwen2_5_omni_token2wav.py         │   │   │   ├── test_qwen2_5_omni_talker.py  ⬜
-│   │   └── qwen3_omni/                          │   │   │   └── test_qwen2_5_omni_token2wav.py  ⬜
-│   │       └── ...                               │   │   └── qwen3_omni/
-│   ├── stage_configs/                           │   │       └── test_*.py              ⬜
-│   │   └── *.yaml                               │   └── stage_configs/                 (used by e2e, test_*.py can be added)  ⬜
-│   └── stage_input_processors/                  │   └── stage_input_processors/
-│       └── ...                                  │       └── test_*.py                 ⬜
-│
-├── sample/                             →     ├── sample/
-│   └── __init__.py                             │   └── test_*.py                      ⬜
-│
-├── utils/                              →     ├── utils/
-│   └── __init__.py                             │   └── test_*.py                       ⬜  (no platform_utils.py currently)
-│
-├── worker/                             →     ├── worker/
-│   ├── gpu_ar_model_runner.py                  │   ├── test_gpu_ar_model_runner.py    ⬜
-│   ├── gpu_ar_worker.py                        │   ├── test_gpu_ar_worker.py           ⬜
-│   ├── gpu_generation_model_runner.py          │   ├── test_gpu_generation_model_runner.py  ✅
-│   ├── gpu_generation_worker.py                │   ├── test_gpu_generation_worker.py  ⬜
-│   ├── gpu_model_runner.py                     │   ├── test_omni_gpu_model_runner.py   ✅
-│   └── mixins.py                               │   └── (npu under platforms/npu/worker/)  # not worker/npu/
-│
-├── platforms/                          →     (no tests/platforms/, e2e and stage_configs provide indirect coverage)
-│   ├── cuda/
-│   ├── npu/worker/                             # NPU worker here, not vllm_omni/worker/npu/
-│   ├── rocm/
-│   └── xpu/worker/
-│
-├── outputs.py                          →     test_outputs.py                         ✅ (at tests root)
-├── (logger, patch, request, version)    →     (no corresponding unit test)
-│
-└── e2e (tests side only)               →     ├── e2e/
-                                               ├── online_serving/                     ✅ non-empty
-                                               │   ├── test_qwen2_5_omni.py
-                                               │   ├── test_async_omni.py
-                                               │   ├── test_qwen3_omni.py
-                                               │   ├── test_qwen3_omni_expansion.py
-                                               │   ├── test_mimo_audio.py
-                                               │   └── test_images_generations_lora.py
-                                               └── offline_inference/                  ✅
-                                                   ├── test_qwen2_5_omni.py
-                                                   ├── test_qwen3_omni.py
-                                                   ├── test_bagel_text2img.py
-                                                   ├── test_z_image.py
-                                                   ├── test_wan22.py
-                                                   ├── test_zimage_tensor_parallel.py
-                                                   ├── test_cache_dit.py
-                                                   ├── test_teacache.py
-                                                   ├── test_stable_audio_expansion.py
-                                                   ├── test_diffusion_cpu_offload.py
-                                                   ├── test_diffusion_layerwise_offload.py
-                                                   ├── test_diffusion_lora.py
-                                                   ├── test_sequence_parallel.py
-                                                   ├── test_qwen_image_edit_expansion.py
-                                                   └── stage_configs/                  (legacy schema, still present
-                                                       ├── bagel_*.yaml                 for unmigrated models)
-                                                       └── npu/, rocm/, etc.
-
-# Migrated models (qwen3_omni_moe, qwen2_5_omni, qwen3_tts) live under
-# vllm_omni/deploy/ instead — see docs/configuration/stage_configs.md.
-examples/                                     tests
-│                                             └── examples
-├── online_serving/                     →         ├── online_serving/
-│   └── {doc_page_title}/README.md                │   └── test_{doc_page_title}.py  ⬜
-└── offline_inference/                  →         └── offline_inference/
-    └── {doc_page_title}/README.md                    └── test_{doc_page_title}.py  ⬜
-```
-
-
-
-### Naming Conventions
-
-- **Unit Tests**: Use `test_<module_name>.py` format. Example: `stage_utils.py` → `test_stage_utils.py`
-
-- **E2E Tests**: Place in `tests/e2e/offline_inference/` or `tests/e2e/online_serving/` with descriptive names. Example: `tests/e2e/offline_inference/test_qwen3_omni.py`, `tests/e2e/offline_inference/test_diffusion_model.py`
-
-- **Expansion Tests**
-
-### Best Practices
-
-1. **Mirror Source Structure**: Test directories should mirror the source code structure
-2. **Test Type Indicators**: Use comments to indicate test types (UT for unit tests, E2E for end-to-end tests)
-3. **Shared Resources**: Place shared test configurations (e.g., CI configs) in appropriate subdirectories
-4. **Consistent Naming**: Follow the `test_*.py` naming convention consistently across all test files
-
-
-## Test codes requirements
-
-### Coding style
-
-1. **File header**: Add SPDX license header to all test files
-2. **Imports**: Pls don't use manual `sys.path` modifications, use standard imports instead.
-3. **Test type differentiation**:
-
-      - Unit tests: Maintain mock style
-      - E2E tests for models: Consider using OmniRunner uniformly, avoid decorators
-
-4. **Documentation**: Add docstrings to all test functions
-5. **Environment variables**: Set uniformly in `conftest.py` or at the top of files
-6. **Type annotations**: Add type annotations to all test function parameters
-7. **Pytest Markers**: Add necessary markers like `@pytest.mark.core_model` and use `@hardware_test` to declare hardware requirements (check detailed in [Markers for Tests](../ci/tests_markers.md)).
-
-### Template
-#### E2E - Online serving
-
-E2E Online tests for Qwen3-Omni model with mix input and audio+text output. Based on `tests/e2e/online_serving/test_qwen3_omni.py`.
-
-```python
-"""
-E2E Online tests for Qwen3-Omni model with mix input and audio+text output.
-"""
-
-import os
-
-os.environ["VLLM_WORKER_MULTIPROC_METHOD"] = "spawn"
-
-import threading
-from pathlib import Path
-
-import openai
-import pytest
-
-from tests.helpers.media import (
-    convert_audio_bytes_to_text,
-    cosine_similarity_text,
-    generate_synthetic_video,
-)
-from tests.helpers.runtime import OmniServer, dummy_messages_from_mix_data
-from tests.helpers.stage_config import get_deploy_config_path, modify_stage_config
-from vllm_omni.platforms import current_omni_platform
-
-# Edit: model name and stage config path
-models = ["Qwen/Qwen3-Omni-30B-A3B-Instruct"]
-
-#If you use the default configuration file, you can directly use the following address.
-def get_default_config():
-    return get_deploy_config_path("ci/qwen3_omni_moe.yaml")
-
-#If you need to modify the configuration file, you can use modify_stage_config.
-def get_chunk_config():
-    path = modify_stage_config(
-        get_default_config(),
-        updates={
-            "async_chunk": True,
-            "stage_args": {
-                0: {
-                    "engine_args.custom_process_next_stage_input_func": "vllm_omni.model_executor.stage_input_processors.qwen3_omni.thinker2talker_async_chunk"
-                },
-                1: {
-                    "engine_args.custom_process_next_stage_input_func": "vllm_omni.model_executor.stage_input_processors.qwen3_omni.talker2code2wav_async_chunk"
-                },
-            },
-        },
-        deletes={"stage_args": {2: ["custom_process_input_func"]}},
-    )
-    return path
-
-stage_configs = [get_default_config(), CHUNK_CONFIG_PATH]
-
-test_params = [(model, stage_config) for model in models for stage_config in stage_configs]
-
-
-#Please use this method to launch the online instance.
-_omni_server_lock = threading.Lock()
-
-@pytest.fixture(scope="module")
-def omni_server(request):
-    """Start vLLM-Omni server as a subprocess. Use module scope for multi-stage init (10-20+ min)."""
-    with _omni_server_lock:
-        model, stage_config_path = request.param
-        with OmniServer(
-            model,
-            ["--stage-configs-path", stage_config_path, "--stage-init-timeout", "120"],
-        ) as server:
-            yield server
-
-
-@pytest.fixture
-def client(omni_server):
-    """OpenAI client for the running vLLM-Omni server."""
-    return openai.OpenAI(
-        base_url=f"http://{omni_server.host}:{omni_server.port}/v1",
-        api_key="EMPTY",
-    )
-
-#Please use function definitions above the test function to define the prompts and other parameters you need.
-def get_system_prompt():
-    return {
-        "role": "system",
-        "content": [
-            {
-                "type": "text",
-                "text": (
-                    "You are Qwen, a virtual human developed by the Qwen Team, "
-                    "Alibaba Group, capable of perceiving auditory and visual inputs, "
-                    "as well as generating text and speech."
-                ),
-            }
-        ],
-    }
-
-...
-
-#Please define test case tags according to the instructions in the marker documentation.
-@pytest.mark.core_model
-@pytest.mark.omni
-@pytest.mark.parametrize("omni_server", test_params, indirect=True)
-def test_mix_to_text_audio_001(client: openai.OpenAI, omni_server, request) -> None:
-    # PLEASE FOLLOW THESE TEMPLATE INSTRUCTIONS:
-    # ============================================================================
-    # TEMPLATE USAGE GUIDE:
-    # 1. Copy this entire function as a starting point for multi-modal tests
-    # 2. Update the test name to reflect your specific test scenario
-    # 3. Modify input/output modalities as needed (see OPTIONS section below)
-    # 4. Adjust assertions based on your expected outcomes
-    # 5. Add custom validation logic for your specific use case
-    # ============================================================================
-
-    #Please list the relevant test points.
-    """
-    Test multi-modal input processing and text/audio output generation via OpenAI API.
-    Deploy Setting: default yaml
-    Input Modal: text + audio + video + image
-    Output Modal: text + audio
-    Input Setting: stream=True
-    Datasets: single request
-    """
-    # SECTION 1: TEST SETUP AND INITIALIZATION
-    # =========================================
-    # INSTRUCTIONS: Initialize test variables and prepare test environment
-    # MODIFY: Add any additional test setup required for your scenario
-    e2e_list = list()
-    # SECTION 2: TEST DATA GENERATION
-    # ================================
-    # INSTRUCTIONS: Generate or load test data for each input modality
-    # MODIFY: Replace synthetic generators with your actual data sources
-    # VIDEO DATA - Generate synthetic video for testing
-    # FORMAT: data:video/mp4;base64,{base64_encoded_video}
-    # PARAMETERS: width, height, duration_frames
-    video_data_url = f"data:video/mp4;base64,{generate_synthetic_video(224, 224, 300)['base64']}"
-    # IMAGE DATA - Generate synthetic image for testing
-    # FORMAT: data:image/jpeg;base64,{base64_encoded_image}
-    # PARAMETERS: width, height
-    image_data_url = f"data:image/jpeg;base64,{generate_synthetic_image(224, 224)['base64']}"
-    # AUDIO DATA - Generate synthetic audio for testing
-    # FORMAT: data:audio/wav;base64,{base64_encoded_audio}
-    # PARAMETERS: duration_seconds, channels
-    audio_data_url = f"data:audio/wav;base64,{generate_synthetic_audio(5, 1)['base64']}"
-
-    # SECTION 3: MESSAGE CONSTRUCTION
-    # ================================
-    # INSTRUCTIONS: Assemble the complete message payload for API request
-    # MODIFY: Add/remove modalities or change prompt structure as needed
-
-    # USAGE: Construct a message containing all input modalities
-    # IMPORTANT: Ensure the message structure matches OpenAI API expectations
-    # CUSTOMIZATION POINTS:
-    #   - system_prompt: Controls the assistant's behavior
-    #   - content_text: The user's text prompt/question
-    #   - *_data_url: URLs for media content (video/image/audio)
-    messages = dummy_messages_from_mix_data(
-        system_prompt=get_system_prompt(),
-        video_data_url=video_data_url,
-        image_data_url=image_data_url,
-        audio_data_url=audio_data_url,
-        content_text=get_prompt("mix"),
-    )
-
-    # SECTION 4: API REQUEST EXECUTION
-    # =================================
-    # INSTRUCTIONS: Make the API call and measure performance
-    # MODIFY: Add timeout, retry logic, or additional parameters
-    start_time = time.perf_counter()
-    chat_completion = client.chat.completions.create(model=omni_server.model, messages=messages, stream=True)
-
-    #Call using your preferred method and obtain the final audio and text outputs.
-    ...
-
-    # SECTION 5: OUTPUT VALIDATION
-    # =============================
-    # INSTRUCTIONS: Verify that outputs meet expected criteria
-    # MODIFY: Adjust validation logic for your specific requirements
-
-    # ASSERTION 1: E2E Validation
-    # PURPOSE: Verify that the E2E latency is less than the baseline.
-    current_e2e = time.perf_counter() - start_time
-    print(f"the request e2e is: {current_e2e}")
-    e2e_list.append(current_e2e)
-
-    print(f"the avg e2e is: {sum(e2e_list) / len(e2e_list)}")
-
-
-
-    # ASSERTION 2: Text Output Validation
-    # PURPOSE: Verify that text output was generated with keyword content
-    assert text_content is not None and len(text_content) >= 2, "No text output is generated"
-    assert any(
-        keyword in text_content.lower() for keyword in ["square", "quadrate", "sphere", "globe", "circle", "round"]
-    ), "The output does not contain any of the keywords."
-
-
-    # ASSERTION 3: Cross-Modal Consistency
-    # PURPOSE: Verify text and audio outputs convey the same information
-    # CUSTOMIZATION: Adjust similarity threshold (0.9) based on accuracy requirements
-    assert audio_data is not None, "No audio output is generated"
-    audio_content = convert_audio_bytes_to_text(audio_data)
-    print(f"text content is: {text_content}")
-    print(f"audio content is: {audio_content}")
-    similarity = cosine_similarity_text(audio_content.lower(), text_content.lower())
-    print(f"similarity is: {similarity}")
-    assert similarity > 0.9, "The audio content is not same as the text"
-```
-
-
-#### E2E - Offline inference
-```python
-# SPDX-License-Identifier: Apache-2.0
-# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
-"""
-Offline E2E smoke test for an omni model (video → audio).
-"""
-
-import os
-from pathlib import Path
-
-import pytest
-from vllm.assets.video import VideoAsset
-
-from tests.helpers.mark import hardware_test
-from ..multi_stages.conftest import OmniRunner
-
-# Optional: set process start method for workers
-os.environ["VLLM_WORKER_MULTIPROC_METHOD"] = "spawn"
-
-models = ["{your model name}"] #Edit here to load your model
-stage_configs = [str(Path(__file__).parent / "stage_configs" / {your model yaml})] #Edit here to load your model yaml
-
-# Create parameter combinations for model and stage config
-test_params = [(model, stage_config) for model in models for stage_config in stage_configs]
-
-# function name: test_{input_modality}_to_{output_modality}
-# modality candidate: text, image, audio, video, mixed_modalities
-@pytest.mark.core_model
-@pytest.mark.omni
-@hardware_test(
-    res={"cuda": "L4", "rocm": "MI325", "npu": "A2"},
-    num_cards=2,
-)
-@pytest.mark.parametrize("test_config", test_params)
-def test_video_to_audio(omni_runner: type[OmniRunner], model: str) -> None:
-    """Offline inference: video input, audio output."""
-    model, stage_config_path = test_config
-    with omni_runner(model, seed=42, stage_configs_path=stage_config_path) as runner:
-        # Prepare inputs
-        video = VideoAsset(name="sample", num_frames=4).np_ndarrays
-
-        outputs = runner.generate_multimodal(
-            prompts="Describe this video briefly.",
-            videos=video,
-        )
-
-        # Minimal assertions: got outputs and at least one audio result
-        assert outputs
-        has_audio = any(o.final_output_type == "audio" for o in outputs)
-        assert has_audio
-```

docs/contributing/model/adding_diffusion_model.md

@@ -685,7 +685,7 @@ For a fair comparison, keep the same **prompt**, **seed**, **resolution**, **num
 
 To ensure project maintainability and sustainable development, please submit test code (unit tests, system tests, or end-to-end tests) alongside their code changes.
 
-For comprehensive testing guidelines and the definition of test levels (L1-L5), please refer to the [Multi-Level Automated Testing System Documentation](../ci/CI_5levels.md). You are at least required to add an L4 *functionality* test described in that document.
+For the definition of test levels (L1-L5), see [Test System Overview](../ci/test_system_overview.md). You are at least required to add an L4 *functionality* test as described in [Test Writing Guide](../ci/test_writing_guide.md#l4-level-testing-full-functionality-performance-and-documentation-testing).
 
 ---
 

docs/contributing/model/adding_omni_model.md

@@ -34,6 +34,7 @@ vllm_omni/model_executor/models/
 └── your_model_name/              # Model directory (e.g., qwen3_omni)
     ├── __init__.py               # Exports main model class
     ├── your_model.py             # Main unified model class
+    ├── pipeline.py               # Pipeline topology registry entry
     ├── your_model_stage1_implementation.py      # Stage 1 implementation (e.g., thinker)
     ├── your_model_stage2_implementation.py      # Stage 2 implementation (e.g., talker)
     └── your_model_stage3_implementation.py      # Stage 3 implementation (e.g., code2wav)
@@ -42,10 +43,14 @@ vllm_omni/model_executor/models/
 vllm_omni/model_executor/stage_input_processors/
 └── your_model_name.py            # Stage transition processors
 
-vllm_omni/model_executor/stage_configs/
-└── your_model_name.yaml          # Stage configuration file
+vllm_omni/deploy/
+└── your_model_name.yaml          # Deployment configuration file
 ```
 
+New in-tree models should register their `PipelineConfig` in
+`vllm_omni/config/pipeline_registry.py`; use `--stage-configs-path` only for
+custom legacy `stage_args` YAMLs.
+
 ## Step-by-Step Implementation
 
 ### Step 1: Create the Model Directory
@@ -535,7 +540,7 @@ Do **not** add a no-suffix `thinker2talker` when a `sync_process_input_func` is
 
 ## Testing
 
-For comprehensive testing guidelines, please refer to the [Test File Structure and Style Guide](../ci/tests_style.md).
+For comprehensive testing guidelines, please refer to the [Test Writing Guide](../ci/test_writing_guide.md).
 
 ## Adding a Model Recipe
 

docs/contributing/model/adding_tts_model.md

@@ -175,15 +175,16 @@ vllm_omni/model_executor/models/
   your_model_name/
     __init__.py
     your_model.py                    # Unified class (stage dispatch)
+    pipeline.py                      # Pipeline topology registry entry
     your_model_ar_stage.py           # Stage 0: AR stage
     your_model_decoder.py            # Stage 1: audio decoder
 
 vllm_omni/model_executor/stage_input_processors/
   your_model_name.py                 # Stage 0 -> Stage 1 transition
 
-vllm_omni/model_executor/stage_configs/
-  your_model_name.yaml               # Batch mode config
-  your_model_name_async_chunk.yaml   # Streaming mode config
+vllm_omni/deploy/
+  your_model_name.yaml               # Batch/deploy config
+  your_model_name_async_chunk.yaml   # Streaming/deploy config
 ```
 
 ### Example placement
@@ -445,6 +446,13 @@ Key configuration fields:
 | `final_output` | Whether this stage produces the final user-facing output |
 | `final_output_type` | Type of final output (`audio`, `text`, etc.) |
 
+!!! note
+    New in-tree models should define frozen topology in `models/<model>/pipeline.py`,
+    register it in `vllm_omni/config/pipeline_registry.py`, and put deployment
+    knobs in `vllm_omni/deploy/<model>.yaml`. Legacy `stage_args` YAMLs are still
+    accepted for custom configs via `--stage-configs-path`, but should not be used
+    for new bundled defaults.
+
 ### Batch mode
 
 ```yaml
@@ -629,7 +637,7 @@ Key points:
 
 ## Testing
 
-For general testing conventions, see [tests_style.md](../ci/tests_style.md).
+For general testing conventions, see [Test Writing Guide](../ci/test_writing_guide.md).
 
 Recommended test cases for a new TTS model:
 
@@ -771,8 +779,9 @@ directly from the model's own generator.
 vllm_omni/model_executor/models/your_model_name/
     __init__.py
     modeling_your_model_name.py    # unified class: load_weights + forward + streaming
+    pipeline.py                    # single-stage pipeline topology
 
-vllm_omni/model_executor/stage_configs/your_model_name.yaml
+vllm_omni/deploy/your_model_name.yaml
 ```
 
 No stage input processor is needed.

docs/design/ar_diffusion_pipeline_capability.md

@@ -0,0 +1,165 @@
+# AR-Diffusion pipeline capability
+
+> **Status:** experimental and non-normative. This note documents the current
+> `experimental/ar_diffusion` contract. It follows the ownership and lifecycle
+> direction in [#5137](https://github.com/vllm-project/vllm-omni/issues/5137),
+> but it is not yet part of the stable module-design hierarchy.
+
+`ARDiffusionEngine` is an opt-in diffusion backend for pipelines that keep
+autoregressive attention KV across requests. The engine only selects
+`ARDiffusionModelRunner`; the runner owns paged KV pools, session lookup, LRU
+eviction, and failure cleanup.
+
+A pipeline opts in by implementing `SupportsARDiffusionPipeline`. Its immutable
+`ARDiffusionKVCacheSpec` reports worker-local KV geometry and policy:
+
+- self-attention layers, TP-local KV heads, head size, and tokens per latent frame;
+- latent frames committed by one model block, sliding window, and attention sink;
+- logical KV branches mapped to worker-local storage slots;
+- named fixed-length cross-attention KV, model-specific scratch geometry, and
+  the pipeline's requested session-capacity upper bound.
+
+The runner binds an `ARDiffusionKVState` only around one `forward()` call. The
+pipeline may use it during that context but must not retain it. The session
+survives subsequent requests with the same `session_id`. Request reset, explicit
+close, LRU eviction, and forward exceptions all release runner-owned KV and call
+the pipeline's reset/close lifecycle hook. Cross-attention KV is allocated lazily
+per session and logical KV branch, and is released on the same path. A pipeline
+publishes one named cache transactionally by yielding exactly one `(k, v)` pair
+per layer:
+
+```python
+state.populate_cross_attention("main", "text", project_text_kv_by_layer())
+```
+
+The cache becomes readable only after the iterable completes successfully.
+Logical branches remain independent even when they share a worker-local
+`local_index`; that index only maps lazily executed self-attention and scratch
+work to worker-local capacity slots. It does not define cross-attention cache
+ownership.
+
+Requests select the session with `extra_args["session_id"]`; setting `reset=True`
+replaces it before the forward, while `close_session=True` releases it after a
+successful forward. The runner also exposes `reset_session()` and
+`close_session()` for an owning serving layer.
+
+## Current execution limits
+
+The experimental runner supports request-mode execution with
+`max_num_seqs=1`. It rejects step execution and request-batch execution at load
+time, and its inherited batch/step entry points also fail explicitly. Supporting
+those paths requires a batch-aware state-binding contract rather than silently
+bypassing `bind_ar_diffusion_state()`.
+
+The current implementation retains one session per model instance. A request
+for a different `session_id` closes the resident session through the same LRU
+lifecycle used by explicit close and failure cleanup before allocating the new
+one. Self-attention and cross-attention memory are therefore sized for one
+resident session even if the pipeline declares a larger `session_capacity`.
+This prioritizes rollout throughput and prevents block-pool exhaustion before a
+count-based eviction can run. Multi-session residency can be added later with a
+memory-backed capacity calculation or allocation-pressure eviction.
+
+Scratch storage is also capability-driven. `frames_per_block` reserves space for
+an uncommitted current video block, while `max_scratch_tokens_per_branch`
+declares the maximum model-specific tokens, such as action/state registers, that
+must coexist with it. `AR_DIFFUSION_KV_SCRATCH_BLOCKS_PER_BRANCH` may increase
+that derived minimum for deployment experiments but cannot reduce it.
+
+## DreamZero adapter
+
+DreamZero implements the capability directly on `DreamZeroPipeline`. Its adapter
+reads DreamZero transformer geometry, describes the positive and negative CFG
+branches, binds the state with a context manager, and translates the existing
+boolean CFG calls to the explicit `"positive"` and `"negative"` KV branch names.
+Its optional `ar_diffusion_warmup_requests()` provider owns the robot observation
+schema and the one-frame/four-frame warmup convention; none of those details are
+present in the generic runner.
+
+Code migrating from the former experimental KV state API should use the
+mechanical mapping below:
+
+```python
+# Before: state.pos, state.neg, state.get_kv_caches(is_negative, ...)
+positive = state.adapter("positive")
+negative = state.adapter("negative")
+contexts = state.get_kv_caches("negative", seq_len=frame_tokens, commit_current=True)
+```
+
+Direct experimental `ARDiffusionKVCache` construction also migrates
+mechanically: replace `local_branches` with explicit `kv_branches`,
+`num_frame_per_block` with `frames_per_block`, and the fixed text/image length
+arguments with a named `cross_attention_lengths` mapping plus
+`session_capacity`. Normal engine users do not construct this object directly.
+
+Cross-attention writers migrate from incremental publication:
+
+```python
+# Before
+for layer_index, (key, value) in enumerate(project_text_kv_by_layer()):
+    state.write_cross_attention_kv("text", layer_index, branch, key, value)
+state.mark_cross_attention_populated(branch, "text")
+
+# After
+state.populate_cross_attention(branch, "text", project_text_kv_by_layer())
+```
+
+The iterable is consumed once and must yield exactly `num_layers` pairs. Failed
+or incomplete projection leaves the previous complete cache unchanged, or
+leaves the cache absent on its first population.
+
+## Single-KV-branch pipeline sketch
+
+A future single-branch causal DMD pipeline can opt in without changes to the
+runner or KV core:
+
+```python
+class WorldPipeline:
+    def ar_diffusion_kv_cache_spec(self):
+        return ARDiffusionKVCacheSpec(
+            num_layers=self.transformer.num_layers,
+            num_kv_heads=self.transformer.tp_local_kv_heads,
+            head_size=self.transformer.head_size,
+            tokens_per_frame=self.transformer.tokens_per_latent_frame,
+            frames_per_block=3,
+            window_frames=self.transformer.sliding_window_frames,
+            sink_frames=self.transformer.sink_frames,
+            kv_branches=(ARDiffusionKVBranchSpec("main", 0),),
+            cross_attention=(ARDiffusionCrossAttentionKVSpec("text", self.text_length),),
+            max_scratch_tokens_per_branch=self.transformer.max_register_tokens,
+            session_capacity=1,
+        )
+
+    @contextmanager
+    def bind_ar_diffusion_state(self, session_id, state):
+        self.current_ar_state = state
+        try:
+            yield
+        finally:
+            self.current_ar_state = None
+
+    def reset_ar_diffusion_session(self, session_id):
+        self.model_states.pop(session_id, None)
+
+    def close_ar_diffusion_session(self, session_id):
+        self.model_states.pop(session_id, None)
+```
+
+The model writes/reads named cross-attention KV through the bound state and uses
+the `"main"` KV branch for paged self-attention. Pipelines without a warmup
+provider are loaded normally and skip AR rollout warmup.
+
+## Relationship to the multi-stage KV manager RFC
+
+[#5244](https://github.com/vllm-project/vllm-omni/issues/5244) describes a
+future multi-stage KV transfer manager, connector integration, and cache-aware
+scheduling. This capability has a narrower scope: it owns model-local paged KV
+and session lifecycle inside one AR-Diffusion runner. It does not define
+cross-stage transfer, connector behavior, or scheduler-visible cache affinity.
+
+When the experimental runtime migrates toward #5244, the implementation should
+reconcile its current full `KVCacheManager` use with the RFC's proposed
+`BlockPool`-oriented DiT manager. The pipeline capability should remain the
+model-facing geometry and state-binding boundary where possible, while the
+experimental `ARDiffusionKVCache` class itself should not be treated as a stable
+public API.

docs/design/feature/cache_dit.md

@@ -32,10 +32,13 @@ The library supports three main caching strategies:
 
 vLLM-omni integrates cache-dit through the `CacheDiTBackend` class, which provides a unified interface for managing cache-dit acceleration on diffusion models.
 
+Code outside `vllm_omni/diffusion/cache` must import Cache-DiT symbols from the
+package-level API: `vllm_omni.diffusion.cache.cachedit`.
+
 | Method/Class | Purpose | Behavior |
 |--------------|---------|----------|
-| [`CacheDiTBackend`](https://docs.vllm.ai/projects/vllm-omni/en/latest/api/vllm_omni/diffusion/cache/#vllm_omni.diffusion.cache.CacheBackend) | Unified backend interface | Automatically handles enabler selection and cache refresh |
-| [`enable_cache_for_dit()`](https://docs.vllm.ai/projects/vllm-omni/en/latest/api/vllm_omni/diffusion/cache/cache_dit_backend/#vllm_omni.diffusion.cache.cache_dit_backend.enable_cache_for_dit) | Apply caching to transformer | Configures DBCache on transformer blocks |
+| `CacheDiTBackend` | Unified backend interface | Automatically handles enabler selection and cache refresh |
+| `enable_cache_for_dit()` | Apply caching to transformer | Configures DBCache on transformer blocks |
 
 **Key APIs from Cache-DiT:**
 
@@ -47,7 +50,7 @@ vLLM-omni integrates cache-dit through the `CacheDiTBackend` class, which provid
 | `ForwardPattern` | Defines block forward signature patterns: `Pattern_0`, `Pattern_1`, `Pattern_2` |
 | `ParamsModifier` | Per-transformer or per-block-list cache configuration customization |
 | `DBCacheConfig` | Configuration for DBCache parameters (warmup steps, cached steps, thresholds) |
-| `refresh_context()` | Update cache context | Called when `num_inference_steps` changes |
+| `refresh_context()` | Update cache context | Called at every generation boundary |
 
 ---
 
@@ -179,7 +182,8 @@ cache_dit.enable_cache(
 
 ### Registering Custom Implementations
 
-After writing your custom enabler, register it in `CUSTOM_DIT_ENABLERS` in `vllm_omni/diffusion/cache/cache_dit_backend.py`:
+After writing your custom enabler, register it in `CUSTOM_DIT_ENABLERS` in
+`vllm_omni/diffusion/cache/cachedit/model_specific.py`:
 
 ```python
 CUSTOM_DIT_ENABLERS = {
@@ -266,10 +270,10 @@ Complete examples in the codebase:
 
 | Model | Path | Pattern | Notes |
 |-------|------|---------|-------|
-| **Standard DiT** | [`cache_dit_backend.py::enable_cache_for_dit`](https://docs.vllm.ai/projects/vllm-omni/en/latest/api/vllm_omni/diffusion/cache/cache_dit_backend/#vllm_omni.diffusion.cache.cache_dit_backend.enable_cache_for_dit) | Default enabler | Single transformer, automatic |
-| **Wan2.2** | [`cache_dit_backend.py::enable_cache_for_wan22`](https://docs.vllm.ai/projects/vllm-omni/en/latest/api/vllm_omni/diffusion/cache/cache_dit_backend/#vllm_omni.diffusion.cache.cache_dit_backend.enable_cache_for_wan22) | Single or dual-transformer | Auto-detects mode based on transformer_2 presence |
-| **LongCat** | [`cache_dit_backend.py::enable_cache_for_longcat_image`](https://docs.vllm.ai/projects/vllm-omni/en/latest/api/vllm_omni/diffusion/cache/cache_dit_backend/#vllm_omni.diffusion.cache.cache_dit_backend.enable_cache_for_longcat_image) | Multi-block-list | Two block lists in one transformer |
-| **BAGEL** | [`cache_dit_backend.py::enable_cache_for_bagel`](https://docs.vllm.ai/projects/vllm-omni/en/latest/api/vllm_omni/diffusion/cache/cache_dit_backend/#vllm_omni.diffusion.cache.cache_dit_backend.enable_cache_for_bagel) | Omni model | Complex architecture |
+| **Standard DiT** | `cachedit.backend::enable_cache_for_dit` | Default enabler | Single transformer, automatic |
+| **Wan2.2** | `cachedit.model_specific::enable_cache_for_wan22` | Single or dual-transformer | Auto-detects mode based on transformer_2 presence |
+| **LongCat** | `cachedit.config::CacheDiTAdapterConfig` | Declarative block adapter | Two block lists in one transformer |
+| **BAGEL** | `cachedit.model_specific::BagelCachedAdapter` | Custom cached adapter | Complex architecture |
 
 ---
 

docs/design/feature/diffusion_continuous_batching.md

@@ -71,7 +71,7 @@ The scheduler derives its batch capacity from `max_num_seqs` through
 `max_num_running_reqs`.
 
 Batch admission is gated by
-[`SamplingParamsKey`](gh-file:vllm_omni/diffusion/sched/interface.py),
+[`StepBatchSamplingParamsKey`](gh-file:vllm_omni/diffusion/sched/interface.py),
 which is built from shape-sensitive and CFG-sensitive sampling fields. This is
 the core correctness rule for batching: requests are only co-batched when they
 share the same denoise tensor contract.
@@ -100,9 +100,10 @@ request-mode prompt semantics, see
 ## Runner
 
 The runner keeps persistent per-request execution state in
-[`DiffusionRequestState`](gh-file:vllm_omni/diffusion/worker/utils.py),
-while the scheduler owns a separate lightweight request state for queueing and
-lifecycle tracking.
+[`StepRequestState`](gh-file:vllm_omni/diffusion/worker/utils.py),
+while the scheduler owns a separate lightweight
+[`SchedulerRequestState`](gh-file:vllm_omni/diffusion/sched/interface.py)
+for queueing and lifecycle tracking.
 
 For each step, the runner builds an
 [`InputBatch`](gh-file:vllm_omni/diffusion/worker/input_batch.py) from the
@@ -138,9 +139,10 @@ compatibility gating and runner-side `StepInputBatch` packing.
 
 ## Current Limitations
 
-- Experimental feature; use `max_num_seqs=1` for the older conservative path.
+- Experimental feature; use `max_num_seqs=1` for the conservative single-request
+  step path.
 - Only native pipelines that already support `step_execution=True`.
-- Only homogeneous batches keyed by `SamplingParamsKey` are supported.
+- Only homogeneous batches keyed by `StepBatchSamplingParamsKey` are supported.
 - `cache_backend`, KV transfer, and other request-mode extras are not wired
   into the batched step-wise path yet.
 - Future work can relax the current same-shape restriction with richer

docs/design/feature/diffusion_request_level_batching.md

@@ -82,7 +82,7 @@ The scheduler derives its capacity from `max_num_seqs` through
 can decide whether admission wait is useful before scheduling a new wave.
 
 Batch compatibility is controlled by
-[`SamplingParamsKey`](gh-file:vllm_omni/diffusion/sched/interface.py). The key
+[`RequestBatchSamplingParamsKey`](gh-file:vllm_omni/diffusion/sched/interface.py). The key
 contains shape-sensitive and guidance-sensitive fields, including output count
 and LoRA identity. Requests with incompatible shapes, CFG settings, output
 counts, LoRA adapters, or LoRA scales are kept in separate batches.
@@ -147,7 +147,7 @@ outputs do not fall back to pickle IPC for tensor payloads.
 
 - Only pipelines that declare the request-batch contract use fused batch
   execution.
-- Batches are homogeneous under `SamplingParamsKey`; heterogeneous resolution or
+- Batches are homogeneous under `RequestBatchSamplingParamsKey`; heterogeneous resolution or
   incompatible guidance settings do not co-batch yet.
 - FIFO scheduling can reduce batching opportunities when an incompatible
   request is at the front of the queue.

docs/design/feature/diffusion_step_execution.md

@@ -32,7 +32,7 @@ Current engine/runtime limitations:
 
 - Continuous batching with `max_num_seqs > 1` is experimental and documented in
   [Continuous Batching for Step-Wise Diffusion](diffusion_continuous_batching.md).
-  Keep `max_num_seqs=1` if you want the older conservative behavior.
+  Keep `max_num_seqs=1` for the conservative single-request step path.
 - `cache_backend` is not supported in step mode.
 - Request-mode extras such as KV transfer are not wired into step mode yet.
 - Unsupported pipelines now fail early during model loading instead of failing on the first request.
@@ -49,7 +49,7 @@ state object:
 
 The state lives in
 [`vllm_omni/diffusion/worker/utils.py`](gh-file:vllm_omni/diffusion/worker/utils.py)
-as `DiffusionRequestState`. Store request-scoped tensors there, or use
+as `StepRequestState`. Store request-scoped tensors there, or use
 `state.extra` for model-specific fields that do not justify extending the core
 dataclass.
 

docs/design/feature/disaggregated_inference.md

@@ -66,8 +66,6 @@ runtime:
   connectors:
     connector_of_shared_memory:
       name: SharedMemoryConnector
-      extra:
-        shm_threshold_bytes: 65536
 ```
 
 Wire stages to connectors:

docs/design/feature/omni_connectors/shared_memory_connector.md

@@ -8,9 +8,7 @@ auto-configured when no explicit connector is specified for an edge.
 ## How It Works
 
 All payloads are serialized and stored in shared memory (`/dev/shm`); the SHM
-segment name is returned in metadata. The configuration exposes a
-`shm_threshold_bytes` field for a future inline-vs-SHM split, but the current
-implementation always uses shared memory regardless of payload size.
+segment name is returned in metadata.
 
 ## Configuration
 
@@ -19,8 +17,6 @@ runtime:
   connectors:
     connector_of_shared_memory:
       name: SharedMemoryConnector
-      extra:
-        shm_threshold_bytes: 65536
 ```
 
 ## Notes
@@ -166,27 +162,10 @@ This path is mainly for older code paths and is not the preferred mode for the c
 
 ### 6. Key Implementation Characteristics
 
-#### 6.1 Threshold Exists, but the Current Code Always Uses SHM
+#### 6.1 All Payloads Use Shared Memory
 
-The class keeps a `shm_threshold_bytes` field and still exposes metrics for inline writes. However, the current implementation uses:
-
-```python
-if True:
-    ...
-```
-
-inside `put()`, which means the current code path always writes to shared memory.
-
-So the design still suggests a future split between:
-
-- small payloads inline
-- large payloads in shared memory
-
-but the current behavior is effectively:
-
-- all payloads go through shared memory
-
-This should be documented because it affects real runtime behavior.
+`put()` writes every serialized payload to shared memory. The connector has no
+inline-payload path or size threshold.
 
 #### 6.2 Cleanup Is Currently Passive
 

docs/design/feature/ray_based_execution.md

@@ -47,7 +47,7 @@ When running on Ray, the system automatically adapts its communication strategy:
 
 *   **Cross-Node**: Recommended to use `MooncakeTransferEngineConnector` (RDMA, fastest) or `MooncakeStoreConnector` (TCP fallback).
 *   **Same-Node**: Can still use `SharedMemoryConnector` for efficiency, or Ray's native object store (plasma).
-*   **SHM threshold default differs**: when `worker_backend="ray"`, the SharedMemoryConnector default threshold is set to `sys.maxsize`, which forces payloads to go inline (no SHM). Override `shm_threshold_bytes` in the connector config if you want SHM for Ray runs.
+*   **SHM payload routing**: `SharedMemoryConnector` always stores payloads in `/dev/shm`.
 
 ### 2.4 Internal Helpers
 

docs/design/feature/skip_softmax.md

@@ -0,0 +1,88 @@
+# Skip-Softmax
+
+Skip-Softmax is the sparse-attention mode of the `TRTLLM_ATTN` backend. Usage — the config keys
+and how to pick an operating point — is in
+[Diffusion Attention Backends](../../user_guide/diffusion/attention_backends.md). This page explains
+the algorithm.
+
+## Motivation
+
+In a long attention row (a video DiT can have tens of thousands of keys), the softmax weight
+concentrates on a small fraction of the keys; the rest receive near-zero weight and barely move the
+output. Computing softmax and the value-weighted sum over those keys is wasted work. Skip-Softmax
+detects, per block of keys, when a block cannot matter and skips its softmax and its value
+multiply.
+
+It is approximate: a skipped block still carries a small non-zero contribution, so the mode is
+opt-in and off by default.
+
+## The online-softmax pass
+
+Attention is computed in a single streaming pass over the keys, in tiles of 128. Per query row the
+kernel maintains three running values:
+
+- `m` — the largest score seen so far,
+- `l` — the running denominator `Σ exp(sⱼ − m)`,
+- `O` — the running numerator `Σ exp(sⱼ − m)·vⱼ`,
+
+and returns `O / l` at the end. For each key tile it computes the tile's scores `Q · K_jᵀ · scale`,
+updates `m`, and accumulates that tile's contribution into `l` and `O`. Rescaling `l` and `O` when
+`m` grows keeps the single pass numerically exact.
+
+## The skip test
+
+Once a tile's scores are known, its largest score `tile_max` is compared against the running
+maximum:
+
+```text
+if exp(tile_max − running_max) < threshold:
+    skip this tile          # do not compute its softmax weights or its P·V accumulation
+```
+
+`exp(tile_max − running_max)` is an upper bound on the softmax weight any key in the tile can
+receive: if even the tile's best key is far below the current maximum, every key in the tile is
+negligible, and both the softmax (the exponentials) and the `P·V` accumulation for that tile can be
+dropped. The tile's contribution to `l` and `O` is simply skipped.
+
+## What this bounds
+
+Two properties of the test shape the achievable speedup:
+
+- **`Q · K_jᵀ` always runs.** The test needs `tile_max`, which comes from the tile's scores, so the
+  score matmul is never skipped — only the softmax and the `P·V` accumulation are. The score matmul
+  and the value matmul are comparable in cost, so skipping every eligible tile removes at most
+  roughly half the attention work; the kernel-level speedup is bounded well under 2×, not unbounded.
+
+- **The decision is per tile, not per key.** A tile is skipped only when *all* of its keys are
+  collectively negligible; a single important key keeps the whole tile. How many tiles actually
+  qualify depends on the data and rounds down to tile granularity, so `target_sparsity` selects an
+  operating point on a calibrated curve — it is not a promise that a fixed fraction of tiles is
+  skipped.
+
+## The threshold
+
+The per-tile threshold is normalized by sequence length:
+
+```text
+threshold = factor / seqlen
+```
+
+`factor` comes from one of two sources:
+
+- **Calibrated** — `factor = a · exp(b · target_sparsity)`, where `a` and `b` are fit per model (and
+  per expert, for a multi-expert model) so that a given `target_sparsity` lands near that fraction of
+  skipped tiles on the calibration data.
+- **Direct** — `factor = skip_softmax.threshold · seqlen`, the calibration-free path; the user sets
+  the threshold themselves.
+
+## Timestep gating
+
+`disabled_until_timestep = D` keeps the mode off during the early, high-noise denoise steps and
+turns it on once the normalized timestep `t` drops to `t ≤ D` (`t` runs `1.0` → `0.0` over the
+schedule). The early steps set the global structure of the output and their errors propagate through
+every later step, so keeping them dense costs a few skipped-tile opportunities but protects fidelity.
+
+`t` is the scheduler's own timestep divided by `num_train_timesteps`, published by the pipeline via
+`DenoiseProgressMixin.record_denoise_step`. It is deliberately not derived from the step index,
+because schedulers space their steps non-uniformly. A pipeline that does not publish a timestep
+stays dense when `disabled_until_timestep` is set.

docs/design/feature/vae_parallel.md

@@ -445,6 +445,7 @@ Complete examples in the codebase:
 | **Z-Image** | `vllm_omni/diffusion/distributed/autoencoders/autoencoder_kl.py` | ✅ | ❌ |
 | **Wan2.2** | `vllm_omni/diffusion/distributed/autoencoders/autoencoder_kl_wan.py` | ✅ | ✅ |
 | **Qwen-Image** | `vllm_omni/diffusion/distributed/autoencoders/autoencoder_kl_qwenimage.py` | ✅ | ❌ |
+| **FLUX.2-dev** | `vllm_omni/diffusion/distributed/autoencoders/autoencoder_kl_flux2.py` | ✅ | ✅ |
 
 ---
 

docs/design/module/dit_module.md

@@ -52,7 +52,7 @@ The diffusion module follows a **multi-process, distributed architecture** with
 
 ### Responsibilities
 
-The `DiffusionEngine` is the **orchestrator** of the diffusion inference system. It manages the lifecycle of worker processes and coordinates the execution flow.
+The `DiffusionEngine` is the **orchestrator** of the diffusion inference system. It owns request admission, scheduler/executor coordination, output-stream delivery, cancellation cleanup, and startup warmup.
 
 ### Key Components
 
@@ -60,68 +60,84 @@ The `DiffusionEngine` is the **orchestrator** of the diffusion inference system.
 
 ```python
 class DiffusionEngine:
-    def __init__(self, od_config: OmniDiffusionConfig):
+    def __init__(
+        self,
+        od_config: OmniDiffusionConfig,
+        scheduler: SchedulerInterface | None = None,
+    ):
         self.od_config = od_config
-        self.post_process_func = get_diffusion_post_process_func(od_config)
-        self.pre_process_func = get_diffusion_pre_process_func(od_config)
-        self._processes: list[mp.Process] = []
-        self._make_client()
+        self._init_process_hooks(od_config)
+        self.execution_mode = self._resolve_execution_mode(od_config)
+        self._init_executor(od_config)
+        self._init_scheduler(od_config, scheduler)
+        self._init_runtime_state()
+        self._init_execute_fn()
 ```
 
 **Key Features**:
 
 - **Pre/Post Processing**: Registers model-specific pre-processing and post-processing functions via registry pattern
 
-- **Worker Management**: Launches and manages multiple worker processes (one per GPU)
+- **Execution Mode Selection**: Selects one of two execution modes:
+  - `REQUEST_BATCH`: complete requests are scheduled through request-level execution. `max_num_seqs=1` is the serial request path; values above `1` allow compatible requests to use fused `pipeline.forward(batch)` when the pipeline supports it.
+  - `STEP_BATCH`: requests are advanced one denoising step at a time through `prepare_encode()`, `denoise_step()`, `step_scheduler()`, and `post_decode()`. `max_num_seqs` controls the maximum number of compatible active requests in one step wave.
 
-- **Process Isolation**: Uses multiprocessing for true parallelism
+- **Scheduler Selection**: Uses `StepScheduler` for step-batch execution and `RequestScheduler` for request-batch execution. Tests and custom integrations may inject a `SchedulerInterface` instance explicitly.
 
-#### 1.2 Worker Launch Process
+- **Executor Selection**: Builds the configured `DiffusionExecutor`, then routes scheduled work through `execute_batch` or `execute_step` according to the execution mode. `execute_batch` handles the serial `max_num_seqs=1` request path internally.
 
-The engine launches workers using a **spawn** method:
+- **Factory-Owned Warmup**: `DiffusionEngine.make_engine()` constructs the selected engine backend and then runs startup warmup through `run_startup_warmup()`.
+
+#### 1.2 Execution Modes
 
 ```python
-def _launch_workers(self, broadcast_handle):
-    # Creates one process per GPU
-    for i in range(num_gpus):
-        process = mp.Process(
-            target=worker_proc.worker_main,
-            args=(i, od_config, writer, broadcast_handle),
-            name=f"DiffusionWorker-{i}",
-        )
-        process.start()
+class DiffusionExecutionMode(str, Enum):
+    REQUEST_BATCH = "request_batch"
+    STEP_BATCH = "step_batch"
+```
+
+`streaming_output=True` requires step execution. If streaming is enabled without `step_execution=True`, the engine enables step execution so the output stream can receive intermediate chunks or the final step output.
+
+#### 1.3 Output Stream Lifecycle
+
+```python
+request_id = engine.add_request(request)
+async for output in engine.get_output_stream(request_id):
+    yield output
 ```
 
 **Design Decisions**:
 
-- **Spawn Method**: Ensures clean state for each worker (no shared memory issues)
+- **Unified Stream Semantics**: `step_streaming()` is the public async engine path. Streaming callers forward every yielded output. Non-streaming callers consume the same stream to completion and return only the final output.
 
-- **Pipe Communication**: Uses `mp.Pipe` for initialization handshake
+- **Queue Delivery**: Each admitted request gets an `asyncio.Queue[DiffusionOutput]`. The busy loop emits chunks/final outputs into that queue.
 
-- **Device Selection**: Each worker is assigned a specific GPU (`cuda:{rank}`)
+- **Cancellation Cleanup**: If a client disconnects or cancels stream consumption, `get_output_stream()` removes only the queue. Scheduler terminal state is still finalized when the scheduler reports the request as finished, so request IDs do not remain active indefinitely.
 
-#### 1.3 Request Processing Flow
+#### 1.4 Request Processing Flow
 
 ```python
-def step(self, requests: list[OmniDiffusionRequest]):
+async def step_streaming(self, request: OmniDiffusionRequest):
     # 1. Pre-process requests
-    requests = self.pre_process_func(requests)
+    request = self.pre_process_func(request)
 
-    # 2. Send to scheduler and wait for response
-    output = self.add_req_and_wait_for_response(requests)
-
-    # 3. Post-process results
-    result = self.post_process_func(output.output)
-    return result
+    # 2. Add request and consume the unified output stream
+    async for output in self.async_add_req_and_stream_response(request):
+        # 3. Post-process each delivered output
+        yield self.postprocess_output(request, output, ...)
 ```
 
 **Flow**:
 
 1. **Pre-processing**: Applies model-specific transformations
 
-2. **Scheduling**: Delegates to scheduler for distribution
+2. **Scheduling**: Delegates request state transitions to the selected scheduler
 
-3. **Post-processing**: Converts raw outputs to final format (e.g., PIL images)
+3. **Execution**: Calls the selected executor function for request, request-batch, or step execution
+
+4. **Output Delivery**: Emits intermediate chunks and final outputs through the request output stream
+
+5. **Post-processing**: Converts raw outputs to final format (e.g., PIL images)
 
 ---
 
@@ -135,28 +151,7 @@ The scheduler is a **request-state scheduler**. It owns request lifecycle manage
 
 ### Key Components
 
-#### 2.1 Scheduler Interface
-
-```python
-class SchedulerInterface(ABC):
-    def add_request(self, request: OmniDiffusionRequest) -> str: ...
-    def schedule(self) -> DiffusionSchedulerOutput: ...
-    def update_from_output(
-        self,
-        sched_output: DiffusionSchedulerOutput,
-        output: DiffusionOutput,
-    ) -> set[str]: ...
-```
-
-**Responsibilities**:
-
-- **Lifecycle contract**: Defines how the engine adds requests, triggers one scheduling cycle, and feeds executor results back.
-
-- **Stable boundary**: `DiffusionSchedulerOutput` is the only scheduling result consumed by `DiffusionEngine`.
-
-- **Pluggability**: Different scheduler policies can reuse the same engine integration path.
-
-#### 2.2 Request State Model
+#### 2.1 Request State Model
 
 ```python
 class DiffusionRequestStatus(enum.IntEnum):
@@ -168,10 +163,12 @@ class DiffusionRequestStatus(enum.IntEnum):
     FINISHED_ERROR = ...
 
 @dataclass
-class DiffusionRequestState:
+class SchedulerRequestState:
     request_id: str
     req: OmniDiffusionRequest
+    sampling_params_key: StepBatchSamplingParamsKey | RequestBatchSamplingParamsKey | None = None
     status: DiffusionRequestStatus = DiffusionRequestStatus.WAITING
+    error: str | None = None
 ```
 
 **Design Features**:
@@ -180,18 +177,34 @@ class DiffusionRequestState:
 
 - **Centralized error handling**: Completion, abort, and error states are all normalized in the scheduler layer.
 
-#### 2.3 Shared Bookkeeping in `_BaseScheduler`
+#### 2.2 Shared Bookkeeping in `BaseScheduler`
 
 ```python
-class _BaseScheduler(SchedulerInterface):
+class BaseScheduler:
     def __init__(self) -> None:
         self._request_states = {}
         self._waiting = deque()
         self._running = []
         self._finished_req_ids = set()
         self.max_num_running_reqs = 1
+
+    def add_request(self, request: OmniDiffusionRequest) -> str: ...
+    def schedule(self) -> DiffusionSchedulerOutput: ...
+    def update_from_output(
+        self,
+        sched_output: DiffusionSchedulerOutput,
+        output: BaseRunnerOutput,
+    ) -> set[str]: ...
 ```
 
+**Responsibilities**:
+
+- **Lifecycle contract**: Defines how the engine adds requests, triggers one scheduling cycle, and feeds executor results back.
+
+- **Stable boundary**: `DiffusionSchedulerOutput` is the only scheduling result consumed by `DiffusionEngine`.
+
+- **Pluggability**: Different scheduler policies can reuse the same engine integration path.
+
 **Design Features**:
 
 - **Common state storage**: Shared request maps and waiting/running sets live in the base class.
@@ -199,12 +212,12 @@ class _BaseScheduler(SchedulerInterface):
 - **Shared cleanup logic**: Duplicate-request checks, finish handling, and state removal are centralized instead of
   duplicated in each policy.
 
-- **Current constraint boundary**: `_BaseScheduler` derives `max_num_running_reqs` from `max_num_seqs`. Request-mode diffusion can use that capacity for compatible independent requests when the configured pipeline declares `supports_request_batch = True`; step-wise diffusion uses the same scheduler capacity for compatible step batches.
+- **Current constraint boundary**: `BaseScheduler` derives `max_num_running_reqs` from `max_num_seqs`. Request-mode diffusion can use that capacity for compatible independent requests when the configured pipeline declares `supports_request_batch = True`; step-wise diffusion uses the same scheduler capacity for compatible step batches.
 
-#### 2.4 Current `RequestScheduler` Policy
+#### 2.3 Current `RequestScheduler` Policy
 
 ```python
-class RequestScheduler(_BaseScheduler):
+class RequestScheduler(BaseScheduler):
     def schedule(self) -> DiffusionSchedulerOutput:
         # 1. keep existing RUNNING requests in the scheduling result
         # 2. pull WAITING requests while capacity remains
@@ -215,11 +228,11 @@ class RequestScheduler(_BaseScheduler):
 
 - **FIFO request scheduling**: Waiting requests are promoted in queue order.
 
-- **Compatible request admission**: `RequestScheduler` admits waiting requests while capacity remains and the request's `SamplingParamsKey` is compatible with the active batch. Request-mode execution keeps each logical request independent, while batch-capable pipelines receive the scheduled requests as a runner-side `DiffusionRequestBatch`.
+- **Compatible request admission**: `RequestScheduler` admits waiting requests while capacity remains and the request's `RequestBatchSamplingParamsKey` is compatible with the active batch. Request-mode execution keeps each logical request independent, while batch-capable pipelines receive the scheduled requests as a runner-side `DiffusionRequestBatch`.
 
 - **Executor result feedback**: `update_from_output()` converts executor output into `FINISHED_COMPLETED` or `FINISHED_ERROR` and returns finished request ids.
 
-#### 2.5 Engine-Driven Execution Loop
+#### 2.4 Engine-Driven Execution Loop
 
 ```python
 request_id = scheduler.add_request(request)
@@ -270,7 +283,7 @@ class WorkerProc:
             self.result_mq = MessageQueue(n_reader=1, ...)
 
         # Initialize GPU worker
-        self.worker = GPUWorker(local_rank=gpu_id, rank=gpu_id, od_config=od_config)
+        self.worker = self._create_worker(gpu_id, od_config, ...)
 ```
 
 **Initialization Steps**:
@@ -292,13 +305,18 @@ class WorkerProc:
 #### 3.2 GPU Worker
 
 ```python
-class GPUWorker:
-    def init_device_and_model(self):
+class DiffusionWorker:
+    def init_device(self):
         # Set distributed environment variables
         os.environ["RANK"] = str(rank)
         os.environ["WORLD_SIZE"] = str(world_size)
 
-        # Initialize PyTorch distributed
+        # Select the device and initialize distributed execution
+        self.device = current_omni_platform.get_torch_device(rank)
+        current_omni_platform.set_device(self.device)
+        self.vllm_config = _create_diffusion_worker_vllm_config(
+            self.device, self.od_config
+        )
         init_distributed_environment(world_size, rank)
         parallel_config = self.od_config.parallel_config
         initialize_model_parallel(
@@ -309,16 +327,18 @@ class GPUWorker:
             pipeline_parallel_size=parallel_config.pipeline_parallel_size,
         )
 
-        # Load model
-        model_loader = DiffusersPipelineLoader(load_config, self.od_config)
-        self.pipeline = model_loader.load_model(load_device=f"cuda:{rank}")
+    def load_model(self, load_format="default", custom_pipeline_name=None):
+        self.model_runner.load_model(
+            load_format=load_format,
+            custom_pipeline_name=custom_pipeline_name,
+        )
 
-        # Setup cache backend
-        from vllm_omni.diffusion.cache.selector import get_cache_backend
-        self.cache_backend = get_cache_backend(od_config.cache_backend, od_config.cache_config)
-
-        if self.cache_backend is not None:
-            self.cache_backend.enable(self.pipeline)
+    def init_lora_manager(self):
+        self.lora_manager = DiffusionLoRAManager(
+            pipeline=self.model_runner.pipeline,
+            device=self.device,
+            dtype=self.od_config.dtype,
+        )
 ```
 
 **Key Features**:
@@ -332,17 +352,17 @@ class GPUWorker:
 #### 3.3 Worker Busy Loop
 
 ```python
-def worker_busy_loop(self):
+def _worker_busy_loop(self):
     while self._running:
         # 1. Receive unified message (generation request, RPC request, or shutdown)
-        msg = self.recv_message()
+        msg = self.mq.dequeue(timeout=1.0)
 
         # 2. Route message based on type
         if isinstance(msg, dict) and msg.get("type") == "rpc":
             # Handle RPC request
-            result, should_reply = self.execute_rpc(msg)
+            result, should_reply = self._execute_rpc(msg)
             if should_reply:
-                self.return_result(result)
+                self._return_result(result)
 
         elif isinstance(msg, dict) and msg.get("type") == "shutdown":
             # Handle shutdown message
@@ -351,7 +371,7 @@ def worker_busy_loop(self):
         else:
             # Handle generation request (OmniDiffusionRequest list)
             output = self.worker.execute_model(msg, self.od_config)
-            self.return_result(output)
+            self._return_result(output)
 ```
 
 **Execution Flow**:
@@ -478,7 +498,7 @@ class Attention(nn.Module):
             role_category=role_category,
         )
         self.attn_impl_cls = attn_backend_cls.get_impl_cls()
-        self.attention = self.attn_impl_cls(..., backend_kwargs=spec.extra if spec else None)
+        self.attention = self.attn_impl_cls(..., backend_kwargs=spec.backend_kwargs() if spec else None)
 ```
 
 **Available Backends**:
@@ -520,7 +540,7 @@ def get_attn_backend_for_role(role, head_size, attention_config=None, role_categ
 
 4. **Platform default** — `current_omni_platform.get_diffusion_attn_backend_cls(...)` picks the best-available kernel for the hardware.
 
-`AttentionSpec.extra` is forwarded to the backend constructor as `backend_kwargs`, so backend-specific parameters (e.g. SparseBlock `block_size`) can be set without changing model code.
+A backend that needs configuration exposes it as a typed field on `AttentionSpec`; `AttentionSpec.backend_kwargs()` serializes it into the `backend_kwargs` the impl consumes. Today the only such field is `skip_softmax` (TRTLLM_ATTN), so e.g. `target_sparsity` can be set without changing model code.
 
 For the user-facing CLI surface, see [Diffusion Attention Backends](../../user_guide/diffusion/attention_backends.md). For the role declaration contract on the model side, see [Adding a Diffusion Model](../../contributing/model/adding_diffusion_model.md).
 
@@ -749,7 +769,7 @@ class TeaCacheBackend(CacheBackend):
 
 **2. Cache-DiT Backend**
 
-**Location**: `vllm_omni/diffusion/cache/cache_dit_backend.py`
+**Location**: `vllm_omni/diffusion/cache/cachedit/`
 
 ```python
 class CacheDiTBackend(CacheBackend):
@@ -911,36 +931,49 @@ def initialize_model_parallel(
 1. User Request
    └─> Omni.generate(prompt)
        └─> Prepare OmniDiffusionRequest
-           └─> DiffusionEngine.step(requests)
+           └─> DiffusionEngine.step_streaming(request)
 
 2. Pre-processing
-   └─> pre_process_func(requests)
+   └─> pre_process_func(request)
        └─> Model-specific transformations
 
 3. Scheduling
-   └─> scheduler.add_request(request)
-       └─> scheduler.schedule()
-           └─> DiffusionEngine submits scheduled request to executor.add_req(req)
+       └─> scheduler.add_request(request)
+           └─> scheduler.schedule()
+           └─> DiffusionEngine selects execute_batch / execute_step
 
 4. Worker Execution
-   └─> WorkerProc.worker_busy_loop()
-       └─> GPUWorker.execute_model(reqs)
-           └─> Pipeline.forward(req)
-               ├─> encode_prompt()
-               ├─> prepare_latents()
-               ├─> diffuse() [loop]
-               │   ├─> transformer.forward() [with cache backend hooks]
-               │   └─> scheduler.step()
-               └─> vae.decode()
+   └─> WorkerProc._worker_busy_loop()
+       └─> DiffusionWorker.execute_model(req)
+           └─> DiffusionModelRunner selects execution mode
+               ├─> REQUEST_BATCH
+               │   └─> Pipeline.forward(req_or_batch)
+               │       ├─> encode_prompt()
+               │       ├─> prepare_latents()
+               │       ├─> diffuse() [loop]
+               │       │   ├─> transformer.forward() [with cache backend hooks]
+               │       │   └─> scheduler.step()
+               │       └─> vae.decode()
+               └─> STEP
+                   ├─> pipeline.prepare_encode(state) [new requests]
+                   ├─> pipeline.denoise_step(input_batch)
+                   ├─> pipeline.step_scheduler(state, noise_pred)
+                   └─> pipeline.post_decode(state) [chunk boundary or final step]
 
 5. Result Collection
-   └─> Executor returns DiffusionOutput
+   └─> Executor returns RunnerOutput / BatchRunnerOutput
        └─> scheduler.update_from_output(...)
-           └─> DiffusionEngine pops finished request state
+           ├─> DiffusionEngine emits chunk/final output to request queue
+           └─> DiffusionEngine finalizes terminal scheduler state
 
 6. Post-processing
    └─> post_process_func(output)
        └─> Convert to PIL images / final format
+
+7. Stream Completion / Cancellation
+   ├─> Streaming callers receive every yielded output
+   ├─> Non-streaming callers drain the stream and return the final output
+   └─> Cancelled streams drop their output queue but still allow scheduler state cleanup
 ```
 
 ---

docs/features/session_state_manager.md

@@ -0,0 +1,160 @@
+# Session State Manager (experimental)
+
+!!! warning
+    This feature is **experimental** (RFC [#4480](https://github.com/vllm-project/vllm-omni/issues/4480)).
+    The code lives under `vllm_omni/experimental/world_models/` and its APIs may
+    change without notice. It is **off by default**; when disabled, model
+    behavior is byte-for-byte unchanged.
+
+Autoregressive diffusion world models (DreamZero, Cosmos3, and similar) keep
+per-session state across `forward()` calls: accumulated video latents, frame
+stitching buffers, encode-once conditioning, counters, and reset heuristics.
+Historically each model hand-rolled this state. The session state manager
+extracts it into a shared, typed contract:
+
+- `StateObject` — one typed unit of session state (e.g. `LatentBuffer` for
+  accumulated latents and bounded rings), with a uniform lifecycle
+  (`allocate` / `commit` / `view` / `reset` / `evict` / `nbytes_by_device`).
+- `SessionStateManager` — maps `session_id -> {name: StateObject}`, caps the
+  number of retained sessions by evicting the least recently used one first
+  (matching the bespoke caches it replaces), and reports per-device byte usage
+  via `stats()` for observability.
+
+Attention KV is *not* managed here: for DreamZero it is owned by the
+AR-Diffusion engine's paged KV pool (PR
+[#4534](https://github.com/vllm-project/vllm-omni/pull/4534)). The manager
+covers the model's non-KV session state; integrating the engine's per-session
+KV handle into the same byte accounting is RFC #4480 Phase 1.
+
+## Enabling
+
+Via config:
+
+```yaml
+# deploy config
+enable_session_state_manager: true
+```
+
+or programmatically with `OmniDiffusionConfig(enable_session_state_manager=True)`.
+
+Via environment variables (no config change needed):
+
+```bash
+export OMNI_DIFFUSION_SESSION_STATE_MANAGER=1            # 1/0/true/false/yes/no/on/off
+export OMNI_DIFFUSION_SESSION_STATE_MANAGER_MAX_SESSIONS=64   # optional, positive int
+```
+
+**Precedence:** a *set* environment variable overrides the config value, in
+both directions — `OMNI_DIFFUSION_SESSION_STATE_MANAGER=0` force-disables the
+manager even if the config enables it. Unset or unparsable values fall back
+to the config (an unparsable `MAX_SESSIONS` logs a warning and is ignored).
+
+**Why both a config field and an environment variable?** The config field is
+the API. The environment override exists for A/B equivalence validation: the
+manager-backed path must be bit-identical to the bespoke path, and the way to
+verify that is to run the *same* deploy config twice with only the flag
+flipped per process — no config-file edits, so the two runs cannot drift
+apart. (Same pattern as the `DIFFUSION_CACHE_BACKEND` fallback.)
+
+**Why is `MAX_SESSIONS` not a config field?** The retained-session cap is not
+a public tuning knob: it mirrors the bespoke `MAX_DREAMZERO_SESSIONS = 64`
+constant (which is not configurable either), so the manager-backed path evicts
+exactly like the path it replaces. The environment variable exists only to
+stress eviction in tests and experiments. It is deliberately not promoted to
+the config, because RFC #4480 Phase 1 replaces count-based capping with a byte
+budget — that budget, not this cap, is the knob that deserves a config field.
+
+## Choosing between an attribute and a `StateObject`
+
+A session holds its values in two places. Named `StateObject`s live in the
+session's object map; everything else lives in `SessionState.attrs`, declared
+one line per value with the `SessionAttr` descriptor. The rule for choosing is:
+
+> A value belongs in a `StateObject` when the **manager** must be able to act on
+> it — release it under budget pressure, accumulate into it, stage and commit
+> it, or rebuild it from a recorded source. It belongs in `attrs` when the
+> manager only needs to **carry** it: the model assigns it wholesale, reads it
+> back, and clears it, and no one but the model can decide its fate.
+
+Byte accounting is independent of that choice. A session reports every byte it
+holds in either bucket, because the number describes what the process is
+holding, not what the manager may do about it.
+
+The consequence is worth stating plainly: **a value left in `attrs` is a value
+the byte-budget planner cannot free.** In Phase 0, which accounts but does not
+enforce, that is legitimate — but it should be a recorded decision rather than
+an accident, which is what the table below records.
+
+Note what the rule does *not* say. "Is it big" is not the criterion, and
+neither is "does it grow". The largest allocation a DreamZero session holds —
+`vae_enc_feat_map`, 603 MiB — is an *attribute*, because nothing the manager
+can do to it is useful: it is replaced wholesale, it has no recompute source,
+and freeing it mid-session breaks the encoder stream. Ending the session is the
+only release available, and that reclaims attributes anyway. Size alone argues
+for counting a value, which the accounting does regardless of bucket; it does
+not argue for a lifecycle.
+
+What would change that answer is a release the manager can perform without
+losing the value — offloading it to host memory and copying it back is
+transparent, needs no recomputation, and suits a large contiguous cache. The
+contract has no such operation: `evict()` frees storage, it does not park it.
+So the bucket follows the operations that exist, and moves if they do.
+
+### Byte accounting
+
+`SessionStateManager.stats()` reports byte totals keyed by device
+(`nbytes:cuda:0`, `nbytes:cpu`) alongside `total_nbytes`. The split is not
+cosmetic: device memory is the contended resource and host memory generally is
+not, so a single figure covering both answers no question. Which pool a budget
+applies to is a question for whoever enforces it, and is not decided here.
+
+Two details make the numbers honest. Bytes are counted per **storage**, not as
+`numel() * element_size()`, so a narrow slice reports the whole allocation it
+keeps alive rather than the part it exposes. And storages are **deduplicated**
+across the whole session, so two views of one buffer — or an alias between an
+attribute and a state object — are counted once.
+
+## Scope and guarantees
+
+- **Opt-in and equivalent.** With the flag off, models use their bespoke state
+  paths unchanged. With the flag on, the manager-backed path is validated
+  bit-identical to the bespoke path (CPU equivalence tests under
+  `tests/dreamzero/test_session_state_equivalence.py`, plus GPU A/B runs).
+- **Session eviction is count-based (LRU).** Evicting a session drops it from
+  the lookup table only; an adapter still holding the session keeps its state
+  (matching bespoke behavior, where the caller holds the state object).
+- **Byte budget is recorded, not enforced.** `SessionStateManager.stats()`
+  reports per-manager, per-device byte totals; budget *enforcement* (and
+  eviction driven by it) is a later phase of RFC #4480.
+
+## Supported models
+
+### DreamZero
+
+Sizes measured on one A100 at the shipped deploy config (`180x320` per camera,
+three cameras stitched to `352x640`, bfloat16), over a 400-step session.
+
+| Value | Bucket | Size | Notes |
+|---|---|---|---|
+| `vae_enc_feat_map` | `attrs` | **603 MiB** | Wan encoder causal-conv cache, 24 entries. Constant from ~step 50. The largest thing a session holds: 64 sessions is 37.7 GiB. Counted, not releasable — see the rule above. |
+| `video_latents_across_time` | `LatentBuffer` | grows | AR video latent chunks, host memory (`.cpu()`), kept for decode. |
+| `stitched_buffer` | `LatentBuffer` | bounded | Stitched pixel frames, host memory, ring of `FRAMES_PER_CHUNK`. |
+| `vae_encoder_out` | `LatentBuffer` | 0.43 MiB | Bounded ring of `num_frame_per_block` latent frames — see below. |
+| `clip_feas`, `ys`, `language`, `prompt_embeds` | `attrs` | small | Conditioning tensors, written once per session and reread. |
+| `vae_pending_body_frames` | `attrs` | ≤2.6 MiB | Raw frames awaiting a 4-frame body chunk; bounded by construction. |
+| `call_count`, `current_start_frame`, `vae_stream_initialized` | `attrs` | — | Counters and flags. |
+
+Attention KV does not appear here: it is engine-owned (PR #4534).
+
+**The bounded VAE encoder ring.** The bespoke state concatenates every encoded
+chunk into one tensor and clears it only on a *session* reset — a window reset
+deliberately keeps it — so it grows for as long as the session lives, at
+0.215 MiB per latent frame (one per four steps). Its only reader,
+`_vae_stream_get_observation_latents`, quantises the whole history and then
+keeps the last `num_frame_per_block` latent frames. Since `quant_conv` is a
+`WanCausalConv3d` of kernel size 1, pointwise in time, those frames do not
+depend on the discarded ones. The manager-backed state therefore retains a ring
+of `num_frame_per_block` frames instead: identical output, bounded memory, and
+the per-step quantisation no longer scales with session length. Asking for more
+frames than the ring retains raises rather than silently returning a shorter
+window. The bespoke path is unchanged.

docs/getting_started/installation/gpu.md

@@ -34,7 +34,7 @@ vLLM-Omni is a Python library that supports the following GPU variants. The libr
 
 ### Pre-built wheels
 
-Note: Pre-built wheels are currently available for vLLM-Omni 0.11.0rc1, 0.12.0rc1, 0.14.0rc1, 0.14.0, 0.16.0, 0.18.0, 0.20.0, 0.21.0, 0.22.0, 0.23.0, 0.24.0, and 0.25.0. If you need a newer unreleased revision, please [build from source](https://docs.vllm.ai/projects/vllm-omni/en/latest/getting_started/installation/gpu/#build-wheel-from-source).
+Note: Pre-built wheels are currently available for vLLM-Omni 0.11.0rc1, 0.12.0rc1, 0.14.0rc1, 0.14.0, 0.16.0, 0.18.0, 0.20.0, 0.21.0, 0.22.0, 0.23.0, 0.24.0, 0.25.0, and 0.26.0. If you need a newer unreleased revision, please [build from source](https://docs.vllm.ai/projects/vllm-omni/en/latest/getting_started/installation/gpu/#build-wheel-from-source).
 
 === "NVIDIA CUDA"
 

docs/getting_started/installation/gpu/cuda.inc.md

@@ -20,7 +20,7 @@ Therefore, it is recommended to install vLLM and vLLM-Omni with a **fresh new**
 
 vLLM-Omni is built based on vLLM. Please install it with command below.
 ```bash
-uv pip install vllm==0.25.0 --torch-backend=auto
+uv pip install vllm==0.26.0 --torch-backend=auto
 ```
 
 #### Installation of vLLM-Omni
@@ -39,13 +39,13 @@ uv pip install 'vllm-omni[demo]'
 # --8<-- [start:build-wheel-from-source]
 
 #### Installation of vLLM
-If you do not need to modify source code of vLLM, you can directly install the stable 0.25.0 release version of the library
+If you do not need to modify source code of vLLM, you can directly install the stable 0.26.0 release version of the library
 
 ```bash
-uv pip install vllm==0.25.0 --torch-backend=auto
+uv pip install vllm==0.26.0 --torch-backend=auto
 ```
 
-The 0.25.0 release of vLLM ships CUDA 13.0-compatible binaries by default. If you need a different CUDA variant or want to reuse an existing PyTorch installation, build vLLM from source instead.
+The 0.26.0 release of vLLM ships CUDA 13.0-compatible binaries by default. If you need a different CUDA variant or want to reuse an existing PyTorch installation, build vLLM from source instead.
 
 #### Installation of vLLM-Omni
 Since vllm-omni is rapidly evolving, it's recommended to install it from source
@@ -66,12 +66,12 @@ If you want to check, modify or debug with source code of vLLM, install the libr
 ```bash
 git clone https://github.com/vllm-project/vllm.git
 cd vllm
-git checkout v0.25.0
+git checkout v0.26.0
 ```
 Set up environment variables to get pre-built wheels. If there are internet problems, just download the whl file manually. And set `VLLM_PRECOMPILED_WHEEL_LOCATION` as your local absolute path of whl file.
 ```bash
-#For CUDA 13.0 (the default for v0.25.0; the wheel filename has no `+cu130` suffix)
-export VLLM_PRECOMPILED_WHEEL_LOCATION=https://github.com/vllm-project/vllm/releases/download/v0.25.0/vllm-0.25.0-cp38-abi3-manylinux_2_28_x86_64.whl
+#For CUDA 13.0 (the default for v0.26.0; the wheel filename has no `+cu130` suffix)
+export VLLM_PRECOMPILED_WHEEL_LOCATION=https://github.com/vllm-project/vllm/releases/download/v0.26.0/vllm-0.26.0-cp38-abi3-manylinux_2_28_x86_64.whl
 ```
 Install vllm with command below (If you have no existing PyTorch).
 ```bash
@@ -102,7 +102,7 @@ docker run --runtime nvidia --gpus 2 \
     --env "HF_TOKEN=$HF_TOKEN" \
     -p 8091:8091 \
     --ipc=host \
-    vllm/vllm-omni:v0.25.0 \
+    vllm/vllm-omni:v0.26.0 \
     vllm serve Qwen/Qwen3-Omni-30B-A3B-Instruct --omni --port 8091
 ```
 

docs/getting_started/installation/gpu/rocm.inc.md

@@ -13,7 +13,7 @@ vLLM-Omni current recommends the steps in under setup through Docker Images.
 
 vLLM-Omni is built based on vLLM. Please install it with command below.
 ```bash
-uv pip install vllm==0.25.0+rocm723 --extra-index-url https://wheels.vllm.ai/rocm/0.25.0/rocm723
+uv pip install vllm==0.26.0+rocm723 --extra-index-url https://wheels.vllm.ai/rocm/0.26.0/rocm723
 ```
 
 #### Installation of vLLM-Omni
@@ -34,13 +34,13 @@ uv pip install onnxruntime-rocm
 # --8<-- [start:build-wheel-from-source]
 
 #### Installation of vLLM
-If you do not need to modify source code of vLLM, you can directly install the stable 0.25.0 release version of the library
+If you do not need to modify source code of vLLM, you can directly install the stable 0.26.0 release version of the library
 
 ```bash
-uv pip install vllm==0.25.0+rocm723 --extra-index-url https://wheels.vllm.ai/rocm/0.25.0/rocm723
+uv pip install vllm==0.26.0+rocm723 --extra-index-url https://wheels.vllm.ai/rocm/0.26.0/rocm723
 ```
 
-The pre-built 0.25.0 vLLM wheel targets ROCm 7.2.3. If you need a different ROCm stack or want to reuse an existing PyTorch installation, build vLLM from source instead.
+The pre-built 0.26.0 vLLM wheel targets ROCm 7.2.3. If you need a different ROCm stack or want to reuse an existing PyTorch installation, build vLLM from source instead.
 
 #### Installation of vLLM-Omni
 Since vllm-omni is rapidly evolving, it's recommended to install it from source
@@ -58,7 +58,7 @@ If you want to check, modify or debug with source code of vLLM, install the libr
 ```bash
 git clone https://github.com/vllm-project/vllm.git
 cd vllm
-git checkout v0.25.0
+git checkout v0.26.0
 python3 -m pip install -r requirements/rocm.txt
 python3 setup.py develop
 ```
@@ -130,7 +130,7 @@ docker run --rm \
   -v ~/.cache/huggingface:/root/.cache/huggingface \
   --env "HF_TOKEN=$HF_TOKEN" \
   -p 8091:8091 \
-  vllm/vllm-omni-rocm:v0.25.0 \
+  vllm/vllm-omni-rocm:v0.26.0 \
   --model Qwen/Qwen3-Omni-30B-A3B-Instruct --omni --port 8091
 ```
 
@@ -149,7 +149,7 @@ docker run --rm -it \
   -v ~/.cache/huggingface:/root/.cache/huggingface \
   --env "HF_TOKEN=$HF_TOKEN" \
   --entrypoint bash \
-  vllm/vllm-omni-rocm:v0.25.0
+  vllm/vllm-omni-rocm:v0.26.0
 ```
 
 # --8<-- [end:pre-built-images]

docs/getting_started/quickstart.md

@@ -19,10 +19,10 @@ uv venv --python 3.12 --seed
 source .venv/bin/activate
 
 # On CUDA
-uv pip install vllm==0.25.0 --torch-backend=auto
+uv pip install vllm==0.26.0 --torch-backend=auto
 
 # On ROCm
-uv pip install vllm==0.25.0+rocm723 --extra-index-url https://wheels.vllm.ai/rocm/0.25.0/rocm723
+uv pip install vllm==0.26.0+rocm723 --extra-index-url https://wheels.vllm.ai/rocm/0.26.0/rocm723
 
 git clone https://github.com/vllm-project/vllm-omni.git
 cd vllm-omni
@@ -34,7 +34,7 @@ For additional installation methods — please see the [installation guide](inst
 !!! note
     It is important to install the same major & minor version of vLLM and vLLM Omni, otherwise things may not work as expected. If the versions are misaligned, you will see a warning when you import vLLM Omni.
 
-    If you are seeing strange behavior with the `vllm` command not handling the `--omni` flag correctly, you most likely have a version mismatch with vLLM < `0.25.0` and vLLM Omni `0.25.0`, as vLLM Omni no longer hijacks the vLLM entrypoint. Updating vLLM should resolve this issue.
+    If you are seeing strange behavior with the `vllm` command not handling the `--omni` flag correctly, you most likely have a version mismatch with vLLM < `0.26.0` and vLLM Omni `0.26.0`, as vLLM Omni no longer hijacks the vLLM entrypoint. Updating vLLM should resolve this issue.
 
 ## Offline Inference
 

docs/models/supported_models.md

@@ -41,24 +41,23 @@ th {
 | `Wan22S2VPipeline` | Wan2.2-S2V | `Wan-AI/Wan2.2-S2V-14B` | ✅︎ | ✅︎ | ✅︎ | ✅︎ |
 | `Wan22VACEPipeline` | Wan2.1-VACE | `Wan-AI/Wan2.1-VACE-1.3B-diffusers`, `Wan-AI/Wan2.1-VACE-14B-diffusers` | ✅︎ | ✅︎ | ✅︎ | ✅︎ |
 | `Wan22VACEPipeline` | Wan2.2-VACE | `Pyros13/Wan2.2-VACE-Fun-A14B-Diffusers` | ✅︎ |   |   |   |
-| `LTX2Pipeline` | LTX-2-T2V | `Lightricks/LTX-2` | ✅︎ | ✅︎ | | |
-| `LTX2ImageToVideoPipeline` | LTX-2-I2V | `Lightricks/LTX-2` | ✅︎ | ✅︎ | | |
-| `LTX2TwoStagesPipeline` | LTX-2-T2V | `rootonchair/LTX-2-19b-distilled` | ✅︎ | ✅︎ | | |
-| `LTX2ImageToVideoTwoStagesPipeline` | LTX-2-I2V | `rootonchair/LTX-2-19b-distilled` | ✅︎ | ✅︎ | | |
-| `LTX23Pipeline` | LTX-2.3-T2V | `dg845/LTX-2.3-Diffusers` | ✅︎ | ✅︎ | | |
-| `LTX23ImageToVideoPipeline` | LTX-2.3-I2V | `dg845/LTX-2.3-Diffusers` | ✅︎ | ✅︎ | | |
+| `LTX2Pipeline` | LTX-2 / LTX-2.3 one-stage T2V and I2V | `Lightricks/LTX-2`, `diffusers/LTX-2.3-Diffusers` | ✅︎ | ✅︎ | | |
+| `LTX2DistilledPipeline` | LTX-2 distilled two-stage T2V and I2V | `rootonchair/LTX-2-19b-distilled` | ✅︎ | ✅︎ | | |
+| `LingBotVideoPipeline` | LingBot-Video dense and MoE T2V | `robbyant/lingbot-video-dense-1.3b`, `robbyant/lingbot-video-moe-30b-a3b` | ✅︎ | | | |
 | `DreamZeroPipeline` | DreamZero-DROID | `GEAR-Dreams/DreamZero-DROID` | ✅︎ | ✅︎ | ✅︎ | ✅︎ |
 | `HeliosPipeline`, `HeliosPyramidPipeline` | Helios | `BestWishYsh/Helios-Base`, `BestWishYsh/Helios-Mid`, `BestWishYsh/Helios-Distilled` | ✅︎ | ✅︎ | ✅︎ | |
 | `MagiHumanPipeline` | MagiHuman | `SII-GAIR/daVinci-MagiHuman-Base-1080p` | ✅︎ | ✅︎ | | |
 | `OvisImagePipeline` | Ovis-Image | `OvisAI/Ovis-Image` | ✅︎ | ✅︎ | | ✅︎ |
 | `LongcatImagePipeline` | LongCat-Image | `meituan-longcat/LongCat-Image` | ✅︎ | ✅︎ | ✅︎ | ✅︎ |
 | `LongCatImageEditPipeline` | LongCat-Image-Edit | `meituan-longcat/LongCat-Image-Edit` | ✅︎ | ✅︎ | ✅︎ | ✅︎ |
+| `BooguImagePipeline` | Boogu-Image | `Boogu/Boogu-Image-0.1-Base`, `Boogu/Boogu-Image-0.1-Edit` | ✅︎ | | | |
 | `StableDiffusionXLPipeline` | Stable-Diffusion-XL | `stabilityai/stable-diffusion-xl-base-1.0` | ✅︎ | ✅︎ | ✅︎ | ✅︎ |
 | `StableDiffusion3Pipeline` | Stable-Diffusion-3 | `stabilityai/stable-diffusion-3.5-medium` | ✅︎ | ✅︎ | | ✅︎ |
 | `CosyVoice3Model` | CosyVoice3 | `FunAudioLLM/Fun-CosyVoice3-0.5B-2512` | ✅︎ | ✅︎ | | ✅︎ |
 | `OmniVoiceModel` | OmniVoice | `k2-fsa/OmniVoice` | ✅︎ | | | |
 | `VoxCPM2TalkerForConditionalGeneration` | VoxCPM2 | `openbmb/VoxCPM2` | ✅︎ | | | |
 | `MammothModa2ForConditionalGeneration` | MammothModa2-Preview | `bytedance-research/MammothModa2-Preview` | ✅︎ | ✅︎ | | |
+| `MammothModa2ForConditionalGeneration` | MammothModa2-Dev (AR-only image understanding) | `bytedance-research/MammothModa2-Dev` | ✅︎ | | | |
 | `Flux2KleinPipeline` | FLUX.2-klein | `black-forest-labs/FLUX.2-klein-4B`, `black-forest-labs/FLUX.2-klein-9B` | ✅︎ | ✅︎ | ✅︎ | ✅︎ |
 | `FluxKontextPipeline` | FLUX.1-Kontext-dev | `black-forest-labs/FLUX.1-Kontext-dev` | ✅︎ | ✅︎ | | |
 | `FluxPipeline` | FLUX.1-dev | `black-forest-labs/FLUX.1-dev` | ✅︎ | ✅︎ | | ✅︎ |
@@ -83,7 +82,7 @@ th {
 | `MiMoAudioModel` | MiMo-Audio-7B-Instruct | `XiaomiMiMo/MiMo-Audio-7B-Instruct` | ✅︎ | ✅︎ | | |
 | `MiMoV2ASRForCausalLM` | MiMo-V2.5-ASR | `XiaomiMiMo/MiMo-V2.5-ASR` | ✅︎ | ✅︎ | | |
 | `Flux2Pipeline` | FLUX.2-dev | `black-forest-labs/FLUX.2-dev` | ✅︎ | ✅︎ | | |
-| `FishSpeechSlowARForConditionalGeneration` | Fish Speech S2 Pro | `fishaudio/s2-pro` | ✅︎ | ✅︎ | | |
+| `FishSpeechSlowARForConditionalGeneration` | Fish Speech S2 Pro | `fishaudio/s2-pro` | ✅︎ | ✅︎ | | ✅︎ |
 | `DreamIDOmniPipeline` | DreamID-Omni | `XuGuo699/DreamID-Omni` | ✅︎ | ✅︎ | | |
 | `SenseNovaU1Pipeline` | SenseNova-U1 (DiT-only) | `SenseNova/SenseNova-U1-8B-MoT` | ✅︎ | | | |
 | `LancePipeline` | Lance | `bytedance-research/Lance` | ✅︎ | | | |

docs/serving/videos_api.md

@@ -73,7 +73,7 @@ curl -L "http://localhost:8091/v1/videos/${video_id}/content" -o output.mp4
 | `audio_reference` | string | null | JSON-encoded audio reference for speech-to-video: `{"audio_url": "..."}` — supports HTTP(s) URLs or base64 data URLs |
 | `width` | integer | model default | Output video width |
 | `height` | integer | model default | Output video height |
-| `num_frames` | integer | model default | Number of generated frames |
+| `num_frames` | integer | 1 | Number of generated frames |
 | `fps` | integer | model default | Output frames per second |
 | `num_inference_steps` | integer | model default | Number of diffusion steps |
 | `guidance_scale` | number | null | CFG guidance scale for the low-noise stage |