[CI] Enroll the remaining Misc and Samplers jobs in automatic sharding (17 jobs)

.buildkite/test_areas/misc.yaml

@@ -63,6 +63,7 @@ steps:
 - label: ":nvidia: (H200 MIG 35GB) V1 Core"
   device: h200_35gb
   key: v1-core
+  automatic_shard: true
   timeout_in_minutes: 45
   source_file_dependencies: &v1-core-kv-metrics-deps
     - vllm/config/
@@ -98,6 +99,7 @@ steps:
     - pytest -v -s -m 'not cpu_test' v1/core
   mirror:
     amd:
+      automatic_shard: true
       label: ":amd: (MI355 DPX) V1 Core"
       dind: false
       device: mi355_dpx
@@ -197,6 +199,7 @@ steps:
 # the connector or its tests change.
 - label: ":amd: (MI355 DPX) UMBP Connector Unit Tests"
   key: amd-umbp-connector-unit-mi355
+  automatic_shard: true
   dind: false
   device: mi355_dpx
   num_devices: 1
@@ -212,7 +215,7 @@ steps:
   - tests/v1/kv_connector/umbp_test_utils.py
   commands:
   - export VLLM_WORKER_MULTIPROC_METHOD=spawn
-  - pytest -v -s -rs v1/kv_connector/unit/test_umbp_*.py
+  - pytest -v -s -rs v1/kv_connector/unit/test_umbp_async_lookup.py v1/kv_connector/unit/test_umbp_core.py v1/kv_connector/unit/test_umbp_embedded.py v1/kv_connector/unit/test_umbp_events.py v1/kv_connector/unit/test_umbp_hybrid.py
 
 # Single-node MoRIIO P/D on dummy weights, WRITE and READ mode: prefill and
 # decode on separate GPUs over xGMI, compared exactly against a direct serve.
@@ -271,6 +274,7 @@ steps:
 - label: ":nvidia: (H200 MIG 35GB) V1 Metrics + LM Eval"
   device: h200_35gb
   key: v1-metrics-lmeval
+  automatic_shard: true
   timeout_in_minutes: 45
   env:
     # GitHub intermittently rejects HTTP/2 upload-pack requests from the H200
@@ -288,6 +292,7 @@ steps:
     - pytest -v -s entrypoints/openai/correctness/test_lmeval.py::test_lm_eval_accuracy_v1_engine
   mirror:
     amd:
+      automatic_shard: true
       label: ":amd: (MI355 DPX) V1 Metrics + LM Eval"
       dind: false
       device: mi355_dpx
@@ -336,6 +341,7 @@ steps:
 
 - label: ":nvidia: (H200 MIG 18GB) Extract Hidden States Integration"
   key: extract-hidden-states-integration
+  automatic_shard: true
   timeout_in_minutes: 20
   device: h200_18gb
   working_dir: "/vllm-workspace"
@@ -350,6 +356,7 @@ steps:
     - pytest -v -s tests/v1/kv_connector/extract_hidden_states_integration
   mirror:
     amd:
+      automatic_shard: true
       label: ":amd: (MI355 DPX) Extract Hidden States Integration Single GPU"
       dind: false
       device: mi355_dpx
@@ -372,6 +379,7 @@ steps:
 
 - label: ":nvidia: (L4) Extract Hidden States Integration"
   key: extract-hidden-states-integration-2-gpus
+  automatic_shard: true
   timeout_in_minutes: 20
   device: l4
   num_devices: 2
@@ -387,6 +395,7 @@ steps:
     - pytest -v -s -m 'distributed' tests/v1/kv_connector/extract_hidden_states_integration
   mirror:
     amd:
+      automatic_shard: true
       label: ":amd: (MI355) Extract Hidden States Integration"
       dind: false
       device: mi355_2
@@ -497,6 +506,7 @@ steps:
 
 - label: ":nvidia: (L4) Metrics, Tracing"
   key: metrics-tracing-2-gpus
+  automatic_shard: true
   timeout_in_minutes: 25
   device: l4
   num_devices: 2
@@ -521,6 +531,7 @@ steps:
   - pytest -v -s tracing
   mirror:
     amd:
+      automatic_shard: true
       label: ":amd: (MI300) Metrics, Tracing"
       dind: false
       device: mi300_2
@@ -1025,6 +1036,7 @@ steps:
 - label: ":nvidia: (H200 MIG 35GB) [Nightly] Acceptance Length (Large Models)"
   device: h200_35gb
   key: acceptance-length-test-large-models
+  automatic_shard: true
   timeout_in_minutes: 20
   gpu: h100
   optional: true
@@ -1038,6 +1050,7 @@ steps:
     - pytest -v -s v1/spec_decode/test_acceptance_length.py -m slow_test
   mirror:
     amd:
+      automatic_shard: true
       label: ":amd: (MI355 DPX) Acceptance Length (Large Models)"
       dind: false
       device: mi355_dpx

.buildkite/test_areas/samplers.yaml

@@ -5,6 +5,7 @@ steps:
 - label: ":nvidia: (H200 MIG 18GB) Samplers"
   device: h200_18gb
   key: samplers-test
+  automatic_shard: true
   timeout_in_minutes: 60
   source_file_dependencies:
   - vllm/model_executor/layers
@@ -19,6 +20,7 @@ steps:
     - VLLM_USE_FLASHINFER_SAMPLER=1 pytest -v -s samplers -k "not test_beam_search_passes_multimodal_data"
   mirror:
     amd:
+      automatic_shard: true
       label: ":amd: (MI250) Samplers"
       dind: false
       device: mi250_1
@@ -41,6 +43,7 @@ steps:
 - label: ":nvidia: (H200 MIG 35GB) Samplers Multimodal Beam Search"
   device: h200_35gb
   key: samplers-multimodal-beam-search
+  automatic_shard: true
   timeout_in_minutes: 30
   source_file_dependencies:
   - vllm/model_executor/layers
@@ -52,6 +55,7 @@ steps:
     - VLLM_USE_FLASHINFER_SAMPLER=1 pytest -v -s samplers/test_beam_search.py::test_beam_search_passes_multimodal_data
   mirror:
     amd:
+      automatic_shard: true
       label: ":amd: (MI355 DPX) Samplers Multimodal Beam Search"
       dind: false
       device: mi355_dpx