Skip to content

GitLab

Unverified Commit 2df2bb27 authored by Andreas Karatzas's avatar Andreas Karatzas Committed by GitHub
Browse files

[ROCm][CI] Removing all blocking labels from MI355 until stable infra (#34879) 


Signed-off-by: default avatarAndreas Karatzas <akaratza@amd.com>
parent f75b61a9
Show whitespace changes
Inline Side-by-side
Showing with 0 additions and 103 deletions
.buildkite/test-amd.yaml
View file @ 2df2bb27
...@@ -1107,18 +1107,6 @@ steps: ...@@ -1107,18 +1107,6 @@ steps:
commands: commands:
- pytest -v -s models/quantization - pytest -v -s models/quantization
# This test is used only in PR development phase to test individual models and should never run on main
- label: Custom Models Test
mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi325_1
# grade: Blocking
optional: true
commands:
- echo 'Testing custom models...'
# PR authors can temporarily add commands below to test individual models
# e.g. pytest -v -s models/encoder_decoder/vision_language/test_mllama.py
# *To avoid merge conflicts, remember to REMOVE (not just comment out) them before merging the PR*
- label: Transformers Nightly Models Test - label: Transformers Nightly Models Test
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi325_1 agent_pool: mi325_1
...@@ -1709,7 +1697,6 @@ steps: ...@@ -1709,7 +1697,6 @@ steps:
# in /vllm/tools/pre_commit/generate_nightly_torch_test.py # in /vllm/tools/pre_commit/generate_nightly_torch_test.py
mirror_hardwares: [amdexperimental, amdproduction, amdtentative] mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
agent_pool: mi355_1 agent_pool: mi355_1
grade: Blocking
soft_fail: true soft_fail: true
source_file_dependencies: source_file_dependencies:
- requirements/nightly_torch_test.txt - requirements/nightly_torch_test.txt
...@@ -1720,7 +1707,6 @@ steps: ...@@ -1720,7 +1707,6 @@ steps:
timeout_in_minutes: 15 timeout_in_minutes: 15
mirror_hardwares: [amdexperimental, amdproduction, amdtentative] mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
agent_pool: mi355_1 agent_pool: mi355_1
grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
- tests/multimodal - tests/multimodal
...@@ -1733,7 +1719,6 @@ steps: ...@@ -1733,7 +1719,6 @@ steps:
timeout_in_minutes: 30 timeout_in_minutes: 30
mirror_hardwares: [amdexperimental, amdproduction, amdtentative] mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
agent_pool: mi355_1 agent_pool: mi355_1
grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
- tests/test_inputs.py - tests/test_inputs.py
...@@ -1763,7 +1748,6 @@ steps: ...@@ -1763,7 +1748,6 @@ steps:
timeout_in_minutes: 20 timeout_in_minutes: 20
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- tests/standalone_tests/python_only_compile.sh - tests/standalone_tests/python_only_compile.sh
- setup.py - setup.py
...@@ -1774,7 +1758,6 @@ steps: ...@@ -1774,7 +1758,6 @@ steps:
timeout_in_minutes: 30 timeout_in_minutes: 30
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
fast_check: true fast_check: true
torch_nightly: true torch_nightly: true
source_file_dependencies: source_file_dependencies:
...@@ -1791,7 +1774,6 @@ steps: ...@@ -1791,7 +1774,6 @@ steps:
- label: Entrypoints Unit Tests # 5min - label: Entrypoints Unit Tests # 5min
mirror_hardwares: [amdexperimental, amdproduction, amdtentative] mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
agent_pool: mi355_1 agent_pool: mi355_1
grade: Blocking
timeout_in_minutes: 10 timeout_in_minutes: 10
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
fast_check: true fast_check: true
...@@ -1806,7 +1788,6 @@ steps: ...@@ -1806,7 +1788,6 @@ steps:
timeout_in_minutes: 40 timeout_in_minutes: 40
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
fast_check: true fast_check: true
torch_nightly: true torch_nightly: true
...@@ -1824,7 +1805,6 @@ steps: ...@@ -1824,7 +1805,6 @@ steps:
timeout_in_minutes: 130 timeout_in_minutes: 130
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
fast_check: true fast_check: true
torch_nightly: true torch_nightly: true
...@@ -1841,7 +1821,6 @@ steps: ...@@ -1841,7 +1821,6 @@ steps:
timeout_in_minutes: 50 timeout_in_minutes: 50
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
fast_check: true fast_check: true
torch_nightly: true torch_nightly: true
...@@ -1860,7 +1839,6 @@ steps: ...@@ -1860,7 +1839,6 @@ steps:
timeout_in_minutes: 50 timeout_in_minutes: 50
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
fast_check: true fast_check: true
torch_nightly: true torch_nightly: true
...@@ -1875,7 +1853,6 @@ steps: ...@@ -1875,7 +1853,6 @@ steps:
timeout_in_minutes: 50 timeout_in_minutes: 50
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
fast_check: true fast_check: true
torch_nightly: true torch_nightly: true
...@@ -1890,7 +1867,6 @@ steps: ...@@ -1890,7 +1867,6 @@ steps:
timeout_in_minutes: 50 timeout_in_minutes: 50
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
num_gpus: 4 num_gpus: 4
source_file_dependencies: source_file_dependencies:
...@@ -1952,7 +1928,6 @@ steps: ...@@ -1952,7 +1928,6 @@ steps:
timeout_in_minutes: 10 timeout_in_minutes: 10
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_8 agent_pool: mi355_8
# grade: Blocking
gpu: h100 gpu: h100
num_gpus: 8 num_gpus: 8
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
...@@ -1973,7 +1948,6 @@ steps: ...@@ -1973,7 +1948,6 @@ steps:
- label: EPLB Algorithm Test # 5min - label: EPLB Algorithm Test # 5min
mirror_hardwares: [amdexperimental, amdproduction, amdtentative] mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
agent_pool: mi355_1 agent_pool: mi355_1
grade: Blocking
timeout_in_minutes: 15 timeout_in_minutes: 15
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
source_file_dependencies: source_file_dependencies:
...@@ -1985,7 +1959,6 @@ steps: ...@@ -1985,7 +1959,6 @@ steps:
- label: EPLB Execution Test # 10min - label: EPLB Execution Test # 10min
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
timeout_in_minutes: 20 timeout_in_minutes: 20
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
num_gpus: 4 num_gpus: 4
...@@ -2000,7 +1973,6 @@ steps: ...@@ -2000,7 +1973,6 @@ steps:
timeout_in_minutes: 20 timeout_in_minutes: 20
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_2 agent_pool: mi355_2
# grade: Blocking
num_gpus: 2 num_gpus: 2
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2020,7 +1992,6 @@ steps: ...@@ -2020,7 +1992,6 @@ steps:
timeout_in_minutes: 20 timeout_in_minutes: 20
mirror_hardwares: [amdexperimental, amdproduction, amdtentative] mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
agent_pool: mi355_1 agent_pool: mi355_1
grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
- tests/test_regression - tests/test_regression
...@@ -2033,7 +2004,6 @@ steps: ...@@ -2033,7 +2004,6 @@ steps:
timeout_in_minutes: 15 timeout_in_minutes: 15
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
- tests/engine - tests/engine
...@@ -2050,7 +2020,6 @@ steps: ...@@ -2050,7 +2020,6 @@ steps:
# The test uses 4 GPUs, but we schedule it on 8-GPU machines for stability. # The test uses 4 GPUs, but we schedule it on 8-GPU machines for stability.
# See discussion here: https://github.com/vllm-project/vllm/pull/31040 # See discussion here: https://github.com/vllm-project/vllm/pull/31040
agent_pool: mi355_8 agent_pool: mi355_8
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
- tests/v1 - tests/v1
...@@ -2064,7 +2033,6 @@ steps: ...@@ -2064,7 +2033,6 @@ steps:
timeout_in_minutes: 50 timeout_in_minutes: 50
mirror_hardwares: [amdexperimental, amdproduction, amdtentative] mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
agent_pool: mi355_1 agent_pool: mi355_1
grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
- tests/v1 - tests/v1
...@@ -2075,7 +2043,6 @@ steps: ...@@ -2075,7 +2043,6 @@ steps:
timeout_in_minutes: 60 timeout_in_minutes: 60
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
- tests/v1 - tests/v1
...@@ -2103,7 +2070,6 @@ steps: ...@@ -2103,7 +2070,6 @@ steps:
- label: V1 Test attention (H100) # 10min - label: V1 Test attention (H100) # 10min
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
timeout_in_minutes: 30 timeout_in_minutes: 30
gpu: h100 gpu: h100
source_file_dependencies: source_file_dependencies:
...@@ -2143,7 +2109,6 @@ steps: ...@@ -2143,7 +2109,6 @@ steps:
- label: V1 Test others (CPU) # 5 mins - label: V1 Test others (CPU) # 5 mins
mirror_hardwares: [amdexperimental, amdproduction, amdtentative] mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
agent_pool: mi355_1 agent_pool: mi355_1
grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
- tests/v1 - tests/v1
...@@ -2161,7 +2126,6 @@ steps: ...@@ -2161,7 +2126,6 @@ steps:
timeout_in_minutes: 45 timeout_in_minutes: 45
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
working_dir: "/vllm-workspace/examples" working_dir: "/vllm-workspace/examples"
source_file_dependencies: source_file_dependencies:
- vllm/entrypoints - vllm/entrypoints
...@@ -2196,7 +2160,6 @@ steps: ...@@ -2196,7 +2160,6 @@ steps:
timeout_in_minutes: 15 timeout_in_minutes: 15
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
- tests/cuda - tests/cuda
...@@ -2208,7 +2171,6 @@ steps: ...@@ -2208,7 +2171,6 @@ steps:
timeout_in_minutes: 75 timeout_in_minutes: 75
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/model_executor/layers - vllm/model_executor/layers
- vllm/sampling_metadata.py - vllm/sampling_metadata.py
...@@ -2221,7 +2183,6 @@ steps: ...@@ -2221,7 +2183,6 @@ steps:
timeout_in_minutes: 30 timeout_in_minutes: 30
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/lora - vllm/lora
- tests/lora - tests/lora
...@@ -2242,7 +2203,6 @@ steps: ...@@ -2242,7 +2203,6 @@ steps:
timeout_in_minutes: 30 timeout_in_minutes: 30
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
torch_nightly: true torch_nightly: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2259,7 +2219,6 @@ steps: ...@@ -2259,7 +2219,6 @@ steps:
timeout_in_minutes: 30 timeout_in_minutes: 30
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
torch_nightly: true torch_nightly: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2305,7 +2264,6 @@ steps: ...@@ -2305,7 +2264,6 @@ steps:
timeout_in_minutes: 75 timeout_in_minutes: 75
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- csrc/ - csrc/
- tests/kernels/core - tests/kernels/core
...@@ -2317,7 +2275,6 @@ steps: ...@@ -2317,7 +2275,6 @@ steps:
timeout_in_minutes: 35 timeout_in_minutes: 35
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- csrc/attention/ - csrc/attention/
- vllm/v1/attention - vllm/v1/attention
...@@ -2332,7 +2289,6 @@ steps: ...@@ -2332,7 +2289,6 @@ steps:
timeout_in_minutes: 90 timeout_in_minutes: 90
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- csrc/quantization/ - csrc/quantization/
- vllm/model_executor/layers/quantization - vllm/model_executor/layers/quantization
...@@ -2345,7 +2301,6 @@ steps: ...@@ -2345,7 +2301,6 @@ steps:
timeout_in_minutes: 60 timeout_in_minutes: 60
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- csrc/quantization/cutlass_w8a8/moe/ - csrc/quantization/cutlass_w8a8/moe/
- csrc/moe/ - csrc/moe/
...@@ -2362,7 +2317,6 @@ steps: ...@@ -2362,7 +2317,6 @@ steps:
timeout_in_minutes: 45 timeout_in_minutes: 45
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- csrc/mamba/ - csrc/mamba/
- tests/kernels/mamba - tests/kernels/mamba
...@@ -2406,7 +2360,6 @@ steps: ...@@ -2406,7 +2360,6 @@ steps:
torch_nightly: true torch_nightly: true
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/engine/arg_utils.py - vllm/engine/arg_utils.py
- vllm/config/model.py - vllm/config/model.py
...@@ -2423,7 +2376,6 @@ steps: ...@@ -2423,7 +2376,6 @@ steps:
timeout_in_minutes: 20 timeout_in_minutes: 20
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
working_dir: "/vllm-workspace/.buildkite" working_dir: "/vllm-workspace/.buildkite"
source_file_dependencies: source_file_dependencies:
- benchmarks/ - benchmarks/
...@@ -2434,7 +2386,6 @@ steps: ...@@ -2434,7 +2386,6 @@ steps:
timeout_in_minutes: 20 timeout_in_minutes: 20
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
- tests/benchmarks/ - tests/benchmarks/
...@@ -2445,7 +2396,6 @@ steps: ...@@ -2445,7 +2396,6 @@ steps:
timeout_in_minutes: 90 timeout_in_minutes: 90
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- csrc/ - csrc/
- vllm/model_executor/layers/quantization - vllm/model_executor/layers/quantization
...@@ -2466,7 +2416,6 @@ steps: ...@@ -2466,7 +2416,6 @@ steps:
timeout_in_minutes: 75 timeout_in_minutes: 75
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- csrc/ - csrc/
- vllm/model_executor/layers/quantization - vllm/model_executor/layers/quantization
...@@ -2478,7 +2427,6 @@ steps: ...@@ -2478,7 +2427,6 @@ steps:
timeout_in_minutes: 15 timeout_in_minutes: 15
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- csrc/ - csrc/
- vllm/entrypoints/openai/ - vllm/entrypoints/openai/
...@@ -2495,7 +2443,6 @@ steps: ...@@ -2495,7 +2443,6 @@ steps:
timeout_in_minutes: 45 timeout_in_minutes: 45
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
torch_nightly: true torch_nightly: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2508,7 +2455,6 @@ steps: ...@@ -2508,7 +2455,6 @@ steps:
timeout_in_minutes: 45 timeout_in_minutes: 45
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
torch_nightly: true torch_nightly: true
source_file_dependencies: source_file_dependencies:
- vllm/model_executor/models/ - vllm/model_executor/models/
...@@ -2528,7 +2474,6 @@ steps: ...@@ -2528,7 +2474,6 @@ steps:
timeout_in_minutes: 45 timeout_in_minutes: 45
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
torch_nightly: true torch_nightly: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2541,7 +2486,6 @@ steps: ...@@ -2541,7 +2486,6 @@ steps:
- label: Basic Models Test (Other CPU) # 5min - label: Basic Models Test (Other CPU) # 5min
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
timeout_in_minutes: 10 timeout_in_minutes: 10
torch_nightly: true torch_nightly: true
source_file_dependencies: source_file_dependencies:
...@@ -2556,7 +2500,6 @@ steps: ...@@ -2556,7 +2500,6 @@ steps:
timeout_in_minutes: 25 timeout_in_minutes: 25
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
torch_nightly: true torch_nightly: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2570,7 +2513,6 @@ steps: ...@@ -2570,7 +2513,6 @@ steps:
timeout_in_minutes: 45 timeout_in_minutes: 45
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
torch_nightly: true torch_nightly: true
source_file_dependencies: source_file_dependencies:
- vllm/model_executor/models/ - vllm/model_executor/models/
...@@ -2591,7 +2533,6 @@ steps: ...@@ -2591,7 +2533,6 @@ steps:
timeout_in_minutes: 75 timeout_in_minutes: 75
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
torch_nightly: true torch_nightly: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2612,7 +2553,6 @@ steps: ...@@ -2612,7 +2553,6 @@ steps:
timeout_in_minutes: 110 timeout_in_minutes: 110
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
optional: true optional: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2628,7 +2568,6 @@ steps: ...@@ -2628,7 +2568,6 @@ steps:
timeout_in_minutes: 110 timeout_in_minutes: 110
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
optional: true optional: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2640,7 +2579,6 @@ steps: ...@@ -2640,7 +2579,6 @@ steps:
timeout_in_minutes: 50 timeout_in_minutes: 50
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
optional: true optional: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2676,7 +2614,6 @@ steps: ...@@ -2676,7 +2614,6 @@ steps:
timeout_in_minutes: 60 timeout_in_minutes: 60
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
- tests/models/multimodal - tests/models/multimodal
...@@ -2688,7 +2625,6 @@ steps: ...@@ -2688,7 +2625,6 @@ steps:
timeout_in_minutes: 100 timeout_in_minutes: 100
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
torch_nightly: true torch_nightly: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2706,7 +2642,6 @@ steps: ...@@ -2706,7 +2642,6 @@ steps:
timeout_in_minutes: 10 timeout_in_minutes: 10
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
working_dir: "/vllm-workspace/.buildkite/lm-eval-harness" working_dir: "/vllm-workspace/.buildkite/lm-eval-harness"
source_file_dependencies: source_file_dependencies:
- vllm/multimodal/ - vllm/multimodal/
...@@ -2721,7 +2656,6 @@ steps: ...@@ -2721,7 +2656,6 @@ steps:
timeout_in_minutes: 120 timeout_in_minutes: 120
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
optional: true optional: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2736,7 +2670,6 @@ steps: ...@@ -2736,7 +2670,6 @@ steps:
timeout_in_minutes: 120 timeout_in_minutes: 120
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
optional: true optional: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2751,7 +2684,6 @@ steps: ...@@ -2751,7 +2684,6 @@ steps:
timeout_in_minutes: 150 timeout_in_minutes: 150
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
optional: true optional: true
source_file_dependencies: source_file_dependencies:
- vllm/ - vllm/
...@@ -2766,29 +2698,15 @@ steps: ...@@ -2766,29 +2698,15 @@ steps:
timeout_in_minutes: 60 timeout_in_minutes: 60
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- vllm/model_executor/layers/quantization - vllm/model_executor/layers/quantization
- tests/models/quantization - tests/models/quantization
commands: commands:
- pytest -v -s models/quantization - pytest -v -s models/quantization
# This test is used only in PR development phase to test individual models and should never run on main
- label: Custom Models Test
mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1
# grade: Blocking
optional: true
commands:
- echo 'Testing custom models...'
# PR authors can temporarily add commands below to test individual models
# e.g. pytest -v -s models/encoder_decoder/vision_language/test_mllama.py
# *To avoid merge conflicts, remember to REMOVE (not just comment out) them before merging the PR*
- label: Transformers Nightly Models Test - label: Transformers Nightly Models Test
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
working_dir: "/vllm-workspace/" working_dir: "/vllm-workspace/"
optional: true optional: true
commands: commands:
...@@ -2927,7 +2845,6 @@ steps: ...@@ -2927,7 +2845,6 @@ steps:
timeout_in_minutes: 20 timeout_in_minutes: 20
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_2 agent_pool: mi355_2
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
num_gpus: 2 num_gpus: 2
source_file_dependencies: source_file_dependencies:
...@@ -2943,7 +2860,6 @@ steps: ...@@ -2943,7 +2860,6 @@ steps:
timeout_in_minutes: 30 timeout_in_minutes: 30
mirror_hardwares: [amdexperimental, amdmultinode] mirror_hardwares: [amdexperimental, amdmultinode]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
num_gpus: 2 num_gpus: 2
num_nodes: 2 num_nodes: 2
...@@ -2970,7 +2886,6 @@ steps: ...@@ -2970,7 +2886,6 @@ steps:
timeout_in_minutes: 90 timeout_in_minutes: 90
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_2 agent_pool: mi355_2
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
num_gpus: 2 num_gpus: 2
source_file_dependencies: source_file_dependencies:
...@@ -3010,7 +2925,6 @@ steps: ...@@ -3010,7 +2925,6 @@ steps:
timeout_in_minutes: 50 timeout_in_minutes: 50
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_2 agent_pool: mi355_2
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
num_gpus: 2 num_gpus: 2
source_file_dependencies: source_file_dependencies:
...@@ -3032,7 +2946,6 @@ steps: ...@@ -3032,7 +2946,6 @@ steps:
timeout_in_minutes: 60 timeout_in_minutes: 60
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_2 agent_pool: mi355_2
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
num_gpus: 2 num_gpus: 2
source_file_dependencies: source_file_dependencies:
...@@ -3066,7 +2979,6 @@ steps: ...@@ -3066,7 +2979,6 @@ steps:
timeout_in_minutes: 60 timeout_in_minutes: 60
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
num_gpus: 4 num_gpus: 4
source_file_dependencies: source_file_dependencies:
...@@ -3083,7 +2995,6 @@ steps: ...@@ -3083,7 +2995,6 @@ steps:
timeout_in_minutes: 30 timeout_in_minutes: 30
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
num_gpus: 4 num_gpus: 4
source_file_dependencies: source_file_dependencies:
- vllm/lora - vllm/lora
...@@ -3108,7 +3019,6 @@ steps: ...@@ -3108,7 +3019,6 @@ steps:
timeout_in_minutes: 45 timeout_in_minutes: 45
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_2 agent_pool: mi355_2
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
num_gpus: 2 num_gpus: 2
optional: true optional: true
...@@ -3121,7 +3031,6 @@ steps: ...@@ -3121,7 +3031,6 @@ steps:
- label: Weight Loading Multiple GPU Test - Large Models # optional - label: Weight Loading Multiple GPU Test - Large Models # optional
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_2 agent_pool: mi355_2
# grade: Blocking
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
num_gpus: 2 num_gpus: 2
optional: true optional: true
...@@ -3134,7 +3043,6 @@ steps: ...@@ -3134,7 +3043,6 @@ steps:
- label: NixlConnector PD accuracy tests (Distributed) # 30min - label: NixlConnector PD accuracy tests (Distributed) # 30min
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
timeout_in_minutes: 30 timeout_in_minutes: 30
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
num_gpus: 4 num_gpus: 4
...@@ -3148,7 +3056,6 @@ steps: ...@@ -3148,7 +3056,6 @@ steps:
- label: DP EP NixlConnector PD accuracy tests (Distributed) # 15min - label: DP EP NixlConnector PD accuracy tests (Distributed) # 15min
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
timeout_in_minutes: 15 timeout_in_minutes: 15
working_dir: "/vllm-workspace/tests" working_dir: "/vllm-workspace/tests"
num_gpus: 4 num_gpus: 4
...@@ -3165,7 +3072,6 @@ steps: ...@@ -3165,7 +3072,6 @@ steps:
- label: Distributed Tests (A100) # optional - label: Distributed Tests (A100) # optional
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
gpu: a100 gpu: a100
optional: true optional: true
num_gpus: 4 num_gpus: 4
...@@ -3188,7 +3094,6 @@ steps: ...@@ -3188,7 +3094,6 @@ steps:
optional: true optional: true
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
num_gpus: 4 num_gpus: 4
working_dir: "/vllm-workspace/.buildkite/lm-eval-harness" working_dir: "/vllm-workspace/.buildkite/lm-eval-harness"
source_file_dependencies: source_file_dependencies:
...@@ -3204,7 +3109,6 @@ steps: ...@@ -3204,7 +3109,6 @@ steps:
optional: true optional: true
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
num_gpus: 4 num_gpus: 4
working_dir: "/vllm-workspace/.buildkite/lm-eval-harness" working_dir: "/vllm-workspace/.buildkite/lm-eval-harness"
source_file_dependencies: source_file_dependencies:
...@@ -3219,7 +3123,6 @@ steps: ...@@ -3219,7 +3123,6 @@ steps:
- label: Distributed Tests (H200) # optional - label: Distributed Tests (H200) # optional
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_2 agent_pool: mi355_2
# grade: Blocking
gpu: h200 gpu: h200
optional: true optional: true
working_dir: "/vllm-workspace/" working_dir: "/vllm-workspace/"
...@@ -3254,7 +3157,6 @@ steps: ...@@ -3254,7 +3157,6 @@ steps:
timeout_in_minutes: 20 timeout_in_minutes: 20
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_1 agent_pool: mi355_1
# grade: Blocking
source_file_dependencies: source_file_dependencies:
- csrc/ - csrc/
- vllm/model_executor/layers/quantization - vllm/model_executor/layers/quantization
...@@ -3264,7 +3166,6 @@ steps: ...@@ -3264,7 +3166,6 @@ steps:
- label: LM Eval Large Models (4 Card) - label: LM Eval Large Models (4 Card)
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
gpu: a100 gpu: a100
optional: true optional: true
num_gpus: 4 num_gpus: 4
...@@ -3304,7 +3205,6 @@ steps: ...@@ -3304,7 +3205,6 @@ steps:
- label: Prime-RL Integration Test # 15min - label: Prime-RL Integration Test # 15min
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_2 agent_pool: mi355_2
# grade: Blocking
timeout_in_minutes: 30 timeout_in_minutes: 30
optional: true optional: true
num_gpus: 2 num_gpus: 2
...@@ -3319,7 +3219,6 @@ steps: ...@@ -3319,7 +3219,6 @@ steps:
- label: DeepSeek V2-Lite Accuracy - label: DeepSeek V2-Lite Accuracy
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
timeout_in_minutes: 60 timeout_in_minutes: 60
gpu: h100 gpu: h100
optional: true optional: true
...@@ -3331,7 +3230,6 @@ steps: ...@@ -3331,7 +3230,6 @@ steps:
- label: Qwen3-30B-A3B-FP8-block Accuracy (H100) - label: Qwen3-30B-A3B-FP8-block Accuracy (H100)
mirror_hardwares: [amdexperimental, amdproduction] mirror_hardwares: [amdexperimental, amdproduction]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
timeout_in_minutes: 60 timeout_in_minutes: 60
gpu: h100 gpu: h100
optional: true optional: true
...@@ -3354,7 +3252,6 @@ steps: ...@@ -3354,7 +3252,6 @@ steps:
timeout_in_minutes: 60 timeout_in_minutes: 60
mirror_hardwares: [amdexperimental] mirror_hardwares: [amdexperimental]
agent_pool: mi355_4 agent_pool: mi355_4
# grade: Blocking
optional: true optional: true
num_gpus: 4 num_gpus: 4
working_dir: "/vllm-workspace" working_dir: "/vllm-workspace"
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment