diff --git a/.buildkite/test-amd.yaml b/.buildkite/test-amd.yaml index 5620bc5f45..2b20e60d61 100644 --- a/.buildkite/test-amd.yaml +++ b/.buildkite/test-amd.yaml @@ -3177,6 +3177,7 @@ steps: dind: false agent_pool: mi355_2 num_gpus: 2 + optional: true working_dir: "/vllm-workspace/" source_file_dependencies: - benchmarks/attention_benchmarks/ @@ -3921,78 +3922,6 @@ steps: - bash ../tools/install_torchcodec_rocm.sh || exit 1 - pytest -s entrypoints/openai/correctness/ -- label: ":amd: (MI355 DPX) Entrypoints Integration (API Server) %N (MI355 suite)" - timeout_in_minutes: 65 - parallelism: 4 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - optional: true - fast_check: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/entrypoints/serve - - tests/entrypoints/scale_out - commands: - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - pytest -v -s entrypoints/serve --ignore=entrypoints/serve/dev/rpc --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB - - if [ "$$BUILDKITE_PARALLEL_JOB" = "1" ]; then PYTHONPATH=/vllm-workspace pytest -v -s entrypoints/serve/dev/rpc; fi - - pytest -v -s entrypoints/scale_out --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB - -- label: ":amd: (MI355 DPX) Entrypoints Integration (OpenAI API completion) (MI355 suite)" # TBD - timeout_in_minutes: 65 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - fast_check: true - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/entrypoints/openai - commands: - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - pytest -v -s entrypoints/openai/completion --ignore=entrypoints/openai/completion/test_tensorizer_entrypoint.py - - pytest -v -s entrypoints/openai --ignore=entrypoints/openai/completion --ignore=entrypoints/openai/chat_completion --ignore=entrypoints/openai/responses --ignore=entrypoints/openai/correctness - -- label: ":amd: (MI355 DPX) Entrypoints Integration (OpenAI API chat_completion) (MI355 suite)" # TBD - timeout_in_minutes: 70 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - fast_check: true - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/entrypoints/openai - commands: - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - pytest -v -s entrypoints/openai/chat_completion - -- label: ":amd: (MI355 DPX) Entrypoints Integration (API Server Generate) (MI355 suite)" # TBD - timeout_in_minutes: 65 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - fast_check: true - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/tool_use - - tests/entrypoints/tool_parsers - - tests/entrypoints/anthropic - - tests/entrypoints/generate - - tests/entrypoints/cohere - commands: - - pytest -v -s tool_use - - pytest -v -s entrypoints/tool_parsers - - pytest -v -s entrypoints/generate - - pytest -v -s entrypoints/anthropic - - pytest -v -s entrypoints/cohere - - label: ":amd: (MI355 DPX) Entrypoints Integration (Speech to Text)" # TBD timeout_in_minutes: 60 mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] @@ -4435,6 +4364,7 @@ steps: dind: false agent_pool: mi355_2 num_gpus: 2 + optional: true working_dir: "/vllm-workspace" source_file_dependencies: - vllm/model_executor/models/ @@ -4898,7 +4828,7 @@ steps: commands: - pytest -v -s kernels/attention --ignore=kernels/attention/test_triton_unified_attention_diffkv.py --ignore=kernels/attention/test_rocm_aiter_mla_decode.py --ignore=kernels/attention/test_rocm_aiter_mla_decode_metadata.py --ignore=kernels/attention/test_rocm_aiter_mla_fp8_support.py --ignore=kernels/attention/test_rocm_aiter_mla_op_registration.py --shard-id=$$BUILDKITE_PARALLEL_JOB --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT -- label: ":amd: (MI355 DPX) MoE Kernels Shard %N (MI355 suite)" # TBD +- label: ":amd: (MI355 DPX) MoE Kernels Shard %N" # TBD timeout_in_minutes: 55 mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] dind: false @@ -5313,20 +5243,6 @@ steps: commands: - pytest -v -s models/language/generation_ppl_test -- label: ":amd: (MI355 DPX) Language Models (Standard) (MI355 suite)" # TBD - timeout_in_minutes: 60 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/models/language - optional: true - commands: - - pip freeze | grep -E 'torch' - - pytest -v -s models/language -m 'core_model and (not slow_test)' - #---------------------------------------------------- mi355 ยท models / multimodal ----------------------------------------------------# - label: ":amd: (MI355 DPX) Multimodal Models (Standard) 1: qwen2" # TBD @@ -5399,20 +5315,6 @@ steps: commands: - pytest -v -s models/multimodal/pooling -m 'not core_model' -- label: ":amd: (MI355 DPX) Multimodal Models (Standard) 1: qwen2 (MI355 suite)" # TBD - timeout_in_minutes: 65 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/models/multimodal - commands: - - pytest -v -s models/multimodal/generation/test_common.py -m core_model -k "qwen2" - - pytest -v -s models/multimodal/generation/test_ultravox.py -m core_model - - label: ":amd: (MI355 DPX) Multimodal Models (Standard) 4: other + whisper" # TBD timeout_in_minutes: 70 mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] @@ -5933,224 +5835,6 @@ steps: - pytest -v -s v1/attention --ignore=v1/attention/test_mla_backends.py --shard-id=$$BUILDKITE_PARALLEL_JOB --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT - VLLM_ROCM_USE_AITER=1 pytest -v -s v1/attention/test_mla_backends.py --shard-id=$$BUILDKITE_PARALLEL_JOB --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT -- label: ":amd: (MI355 DPX) V1 Core (MI355 suite)" - timeout_in_minutes: 60 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/v1/core - - tests/v1/executor - - tests/v1/kv_offload - - tests/v1/worker - - tests/v1/cudagraph - - tests/v1/kv_connector/unit - - tests/v1/metrics - - tests/entrypoints/openai/correctness/test_lmeval.py - - vllm/config/ - - vllm/distributed/ - - vllm/engine/ - - vllm/entrypoints/pooling/ - - vllm/inputs/ - - vllm/lora/ - - vllm/model_executor/ - - vllm/multimodal/ - - vllm/outputs.py - - vllm/platforms/ - - vllm/pooling_params.py - - vllm/profiler/ - - vllm/sampling_params.py - - vllm/tokenizers/ - - vllm/transformers_utils/ - - vllm/utils/ - - vllm/v1/ - - tests/v1/simple_kv_offload - - tests/v1/streaming_input - - tests/v1/ec_connector/unit - commands: - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - pytest -v -s -m 'not cpu_test' v1/core - -- label: ":amd: (MI355 DPX) V1 Executor + Worker (MI355 suite)" - timeout_in_minutes: 60 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - optional: true - working_dir: /vllm-workspace/tests - source_file_dependencies: - - vllm/ - - tests/v1/core - - tests/v1/executor - - tests/v1/kv_offload - - tests/v1/worker - - tests/v1/cudagraph - - tests/v1/kv_connector/unit - - tests/v1/metrics - - tests/entrypoints/openai/correctness/test_lmeval.py - - vllm/config/ - - vllm/distributed/ - - vllm/engine/ - - vllm/entrypoints/pooling/ - - vllm/inputs/ - - vllm/lora/ - - vllm/model_executor/ - - vllm/multimodal/ - - vllm/outputs.py - - vllm/platforms/ - - vllm/pooling_params.py - - vllm/profiler/ - - vllm/sampling_params.py - - vllm/tokenizers/ - - vllm/transformers_utils/ - - vllm/utils/ - - vllm/v1/ - - tests/v1/simple_kv_offload - - tests/v1/streaming_input - - tests/v1/ec_connector/unit - commands: - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - pytest -v -s v1/executor - - pytest -v -s v1/worker - - pytest -v -s v1/streaming_input - -- label: ":amd: (MI355 DPX) V1 KV Offload (MI355 suite)" - timeout_in_minutes: 60 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - optional: true - working_dir: /vllm-workspace/tests - source_file_dependencies: - - vllm/ - - tests/v1/core - - tests/v1/executor - - tests/v1/kv_offload - - tests/v1/worker - - tests/v1/cudagraph - - tests/v1/kv_connector/unit - - tests/v1/metrics - - tests/entrypoints/openai/correctness/test_lmeval.py - - vllm/config/ - - vllm/distributed/ - - vllm/engine/ - - vllm/entrypoints/pooling/ - - vllm/inputs/ - - vllm/lora/ - - vllm/model_executor/ - - vllm/multimodal/ - - vllm/outputs.py - - vllm/platforms/ - - vllm/pooling_params.py - - vllm/profiler/ - - vllm/sampling_params.py - - vllm/tokenizers/ - - vllm/transformers_utils/ - - vllm/utils/ - - vllm/v1/ - - tests/v1/simple_kv_offload - - tests/v1/streaming_input - - tests/v1/ec_connector/unit - commands: - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - bash /vllm-workspace/.buildkite/scripts/install-kv-offload.sh - - pytest -v -s v1/kv_offload - - pytest -v -s v1/simple_kv_offload - -- label: ":amd: (MI355 DPX) V1 KV Connectors Shard %N (MI355 suite)" - timeout_in_minutes: 60 - parallelism: 4 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - optional: true - working_dir: /vllm-workspace/tests - source_file_dependencies: - - vllm/ - - tests/v1/core - - tests/v1/executor - - tests/v1/kv_offload - - tests/v1/worker - - tests/v1/cudagraph - - tests/v1/kv_connector/unit - - tests/v1/metrics - - tests/entrypoints/openai/correctness/test_lmeval.py - - vllm/config/ - - vllm/distributed/ - - vllm/engine/ - - vllm/entrypoints/pooling/ - - vllm/inputs/ - - vllm/lora/ - - vllm/model_executor/ - - vllm/multimodal/ - - vllm/outputs.py - - vllm/platforms/ - - vllm/pooling_params.py - - vllm/profiler/ - - vllm/sampling_params.py - - vllm/tokenizers/ - - vllm/transformers_utils/ - - vllm/utils/ - - vllm/v1/ - - tests/v1/simple_kv_offload - - tests/v1/streaming_input - - tests/v1/ec_connector/unit - commands: - - bash /vllm-workspace/.buildkite/scripts/install-kv-connectors.sh - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - pytest -v -s -m 'not cpu_test' v1/kv_connector/unit --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB - - pytest -v -s -m 'not cpu_test' v1/ec_connector/unit --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB - -- label: ":amd: (MI355 DPX) V1 Metrics + LM Eval (MI355 suite)" - timeout_in_minutes: 65 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - optional: true - working_dir: /vllm-workspace/tests - source_file_dependencies: - - vllm/ - - tests/v1/core - - tests/v1/executor - - tests/v1/kv_offload - - tests/v1/worker - - tests/v1/cudagraph - - tests/v1/kv_connector/unit - - tests/v1/metrics - - tests/entrypoints/openai/correctness/test_lmeval.py - - vllm/config/ - - vllm/distributed/ - - vllm/engine/ - - vllm/entrypoints/pooling/ - - vllm/inputs/ - - vllm/lora/ - - vllm/model_executor/ - - vllm/multimodal/ - - vllm/outputs.py - - vllm/platforms/ - - vllm/pooling_params.py - - vllm/profiler/ - - vllm/sampling_params.py - - vllm/tokenizers/ - - vllm/transformers_utils/ - - vllm/utils/ - - vllm/v1/ - - tests/v1/simple_kv_offload - - tests/v1/streaming_input - - tests/v1/ec_connector/unit - commands: - - export GIT_CONFIG_COUNT=1 - - export GIT_CONFIG_KEY_0=http.version - - export GIT_CONFIG_VALUE_0=HTTP/1.1 - - export GIT_TERMINAL_PROMPT=0 - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - pytest -v -s -m 'not cpu_test' v1/metrics - - pip install -U git+https://github.com/vllm-project/lm-evaluation-harness.git@streaming-api - - pytest -v -s entrypoints/openai/correctness/test_lmeval.py::test_lm_eval_accuracy_v1_engine - - label: ":amd: (MI355 DPX) CUDAGraph" timeout_in_minutes: 55 mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] @@ -6181,68 +5865,6 @@ steps: - pytest -v -s v1/cudagraph/test_encoder_cudagraph.py -- label: ":amd: (MI355 DPX) V1 Sample (MI355 suite)" - timeout_in_minutes: 70 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/v1/sample - - tests/v1/logits_processors - - tests/v1/test_oracle.py - - tests/v1/test_request.py - - tests/v1/test_outputs.py - - vllm/config/ - - vllm/distributed/ - - vllm/engine/ - - vllm/inputs/ - - vllm/logger.py - - vllm/model_executor/ - - vllm/platforms/ - - vllm/sampling_params.py - - vllm/transformers_utils/ - - vllm/utils/ - - vllm/v1/ - commands: - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - pytest -v -s v1/sample - -- label: ":amd: (MI355 DPX) V1 Logits + Oracle (MI355 suite)" - timeout_in_minutes: 70 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] - dind: false - agent_pool: mi355_dpx - optional: true - working_dir: /vllm-workspace/tests - source_file_dependencies: - - vllm/ - - tests/v1/sample - - tests/v1/logits_processors - - tests/v1/test_oracle.py - - tests/v1/test_request.py - - tests/v1/test_outputs.py - - vllm/config/ - - vllm/distributed/ - - vllm/engine/ - - vllm/inputs/ - - vllm/logger.py - - vllm/model_executor/ - - vllm/platforms/ - - vllm/sampling_params.py - - vllm/transformers_utils/ - - vllm/utils/ - - vllm/v1/ - commands: - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - pytest -v -s v1/logits_processors - - pytest -v -s v1/test_oracle.py - - pytest -v -s v1/test_request.py - - pytest -v -s v1/test_outputs.py - - - label: ":amd: (MI355 DPX) V1 Spec Decode" # TBD timeout_in_minutes: 50 mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355]