[Hardware][AMD][CI] Bump timeouts of various test groups on AMD CI (#47195)
Signed-off-by: Matthew Wong <Matthew.Wong2@amd.com>
Matt committed
Jun 30, 2026 at 14:30 UTC
345b28ff2f75a94e3ab91eb5e50f104dd7b933c2
1 file changed
+14
-14
.buildkite/test-amd.yaml
+14
-14
@@ -418,7 +418,7 @@ steps:
418
#----------------------------------------------------- mi300 · basic_correctness -----------------------------------------------------#
419
420
- label: Basic Correctness # TBD
421
- timeout_in_minutes: 50
421
+ timeout_in_minutes: 95
422
mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300]
423
agent_pool: mi300_1
424
fast_check: true
@@ -436,7 +436,7 @@ steps:
436
- pytest -v -s basic_correctness/test_cpu_offload.py
437
438
- label: Distributed Model Tests (2 GPUs) # TBD
439
- timeout_in_minutes: 65
439
+ timeout_in_minutes: 110
440
mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300]
441
agent_pool: mi300_2
442
num_gpus: 2
@@ -678,7 +678,7 @@ steps:
678
- pytest -v -s distributed/test_eplb_spec_decode.py
679
680
- label: Distributed Tests (2xH100-2xMI300) # TBD
681
- timeout_in_minutes: 30
681
+ timeout_in_minutes: 75
682
mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300]
683
agent_pool: mi300_2
684
num_gpus: 2
@@ -1222,7 +1222,7 @@ steps:
1222
#--------------------------------------------------------- mi300 · examples ----------------------------------------------------------#
1223
1224
- label: Examples # TBD
1225
- timeout_in_minutes: 45
1225
+ timeout_in_minutes: 90
1226
mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300]
1227
agent_pool: mi300_1
1228
optional: true
@@ -1258,7 +1258,7 @@ steps:
1258
#---------------------------------------------------------- mi300 · kernels ----------------------------------------------------------#
1259
1260
- label: Kernels Attention Test %N # TBD
1261
- timeout_in_minutes: 55
1261
+ timeout_in_minutes: 100
1262
mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300]
1263
agent_pool: mi300_1
1264
optional: true
@@ -1292,7 +1292,7 @@ steps:
1292
- pytest -v -s kernels/core --ignore=kernels/core/test_minimax_reduce_rms.py kernels/test_concat_mla_q.py kernels/test_top_k_per_row.py
1293
1294
- label: Kernels MoE Test %N # TBD
1295
- timeout_in_minutes: 50
1295
+ timeout_in_minutes: 95
1296
mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300]
1297
agent_pool: mi300_1
1298
optional: true
@@ -1439,7 +1439,7 @@ steps:
1439
- pytest -v -s models/test_initialization.py::test_can_initialize_small_subset
1440
1441
- label: Basic Models Tests (Other) # TBD
1442
- timeout_in_minutes: 45
1442
+ timeout_in_minutes: 90
1443
mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300]
1444
agent_pool: mi300_1
1445
optional: true
@@ -1903,7 +1903,7 @@ steps:
1903
- pytest -v -s v1/e2e/spec_decode -k "draft_model or no_sync or batch_inference"
1904
1905
- label: Spec Decode Eagle # TBD
1906
- timeout_in_minutes: 45
1906
+ timeout_in_minutes: 90
1907
mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300]
1908
agent_pool: mi300_1
1909
optional: true
@@ -2119,7 +2119,7 @@ steps:
2119
- DP_SIZE=2 pytest -v -s entrypoints/openai/test_multi_api_servers.py
2120
2121
- label: Metrics, Tracing (2 GPUs) # TBD
2122
- timeout_in_minutes: 20
2122
+ timeout_in_minutes: 65
2123
mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300]
2124
agent_pool: mi300_2
2125
optional: true
@@ -2272,7 +2272,7 @@ steps:
2272
#------------------------------------------------------ mi300 · weight_loading -------------------------------------------------------#
2273
2274
- label: Weight Loading Multiple GPU # TBD
2275
- timeout_in_minutes: 30
2275
+ timeout_in_minutes: 75
2276
mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300]
2277
agent_pool: mi300_2
2278
num_gpus: 2
@@ -2284,7 +2284,7 @@ steps:
2284
- bash weight_loading/run_model_weight_loading_test.sh -c weight_loading/models-amd.txt
2285
2286
- label: Weight Loading Multiple GPU - Large Models # TBD
2287
- timeout_in_minutes: 30
2287
+ timeout_in_minutes: 75
2288
mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300]
2289
agent_pool: mi300_2
2290
num_gpus: 2
@@ -2831,7 +2831,7 @@ steps:
2831
- pytest -v -s tests/kernels/attention/test_rocm_aiter_mla_decode_metadata.py
2832
2833
- label: Kernels Attention Test %N # TBD
2834
- timeout_in_minutes: 60
2834
+ timeout_in_minutes: 100
2835
mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355]
2836
agent_pool: mi355_1
2837
parallelism: 2
@@ -3178,7 +3178,7 @@ steps:
3178
#------------------------------------------------------ mi355 · weight_loading -------------------------------------------------------#
3179
3180
- label: Weight Loading Multiple GPU # TBD
3181
- timeout_in_minutes: 30
3181
+ timeout_in_minutes: 75
3182
mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355]
3183
agent_pool: mi355_2
3184
num_gpus: 2
@@ -3190,7 +3190,7 @@ steps:
3190
- bash weight_loading/run_model_weight_loading_test.sh -c weight_loading/models-amd.txt
3191
3192
- label: Weight Loading Multiple GPU - Large Models # TBD
3193
- timeout_in_minutes: 30
3193
+ timeout_in_minutes: 75
3194
mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355]
3195
agent_pool: mi355_2
3196
working_dir: "/vllm-workspace/tests"