diff --git a/.buildkite/test-amd.yaml b/.buildkite/test-amd.yaml index 7a9876032d..0b04b36878 100644 --- a/.buildkite/test-amd.yaml +++ b/.buildkite/test-amd.yaml @@ -1094,25 +1094,6 @@ steps: #----------------------------------------------------------- mi300 · evals -----------------------------------------------------------# -- label: ":amd: (MI300) LM Eval Small Models" # TBD - timeout_in_minutes: 45 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_1 - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - csrc/ - - vllm/model_executor/layers/quantization - - vllm/model_executor/models/ - - vllm/model_executor/model_loader/ - - vllm/v1/attention/backends/ - - vllm/v1/attention/selector.py - - vllm/_aiter_ops.py - - vllm/platforms/rocm.py - optional: true - commands: - - pytest -s -v evals/gsm8k/test_gsm8k_correctness.py --config-list-file=configs/models-small.txt - - label: ":amd: (MI300) DeepSeek V2-Lite Prefetch Offload Accuracy" # TBD timeout_in_minutes: 30 mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] @@ -1153,23 +1134,6 @@ steps: commands: - pytest -s -v evals/mrcr/test_mrcr_correctness.py --config-list-file=evals/mrcr/configs/models-small.txt -- label: ":amd: (MI300) LM Eval KV-Offload" # TBD - timeout_in_minutes: 55 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_1 - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/distributed/kv_transfer/kv_connector/v1/offloading/ - - vllm/distributed/kv_transfer/kv_connector/v1/simple_cpu_offload_connector.py - - vllm/v1/kv_offload/ - - vllm/v1/simple_kv_offload/ - - tests/evals/gsm8k/test_gsm8k_offloading.py - - vllm/platforms/rocm.py - commands: - - pytest -s -v evals/gsm8k/test_gsm8k_offloading.py -k "nemotron-h-8b or gemma-4-e4b-it" - - label: ":amd: (MI300) LM Eval KV-Offload Medium" # TBD timeout_in_minutes: 70 mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] @@ -1188,24 +1152,6 @@ steps: commands: - pytest -s -v evals/gsm8k/test_gsm8k_offloading.py -k "qwen3.5-35b or deepseek-v2-lite" -- label: ":amd: (MI300) LM Eval KV-Offload Large" # TBD - timeout_in_minutes: 65 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_4 - num_gpus: 4 - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/distributed/kv_transfer/kv_connector/v1/offloading/ - - vllm/distributed/kv_transfer/kv_connector/v1/simple_cpu_offload_connector.py - - vllm/v1/kv_offload/ - - vllm/v1/simple_kv_offload/ - - tests/evals/gsm8k/test_gsm8k_offloading.py - - vllm/platforms/rocm.py - commands: - - VLLM_ROCM_USE_AITER=1 pytest -s -v evals/gsm8k/test_gsm8k_offloading.py -k "deepseek-v4-flash" - - label: ":amd: (MI300) LM Eval Small Models Harness" # TBD timeout_in_minutes: 35 mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] @@ -1383,52 +1329,6 @@ steps: commands: - VLLM_ENGINE_READY_TIMEOUT_S=1800 bash .buildkite/scripts/scheduled_integration_test/qwen3_next_mtp_async_eplb.sh 0.8 1319 8040 -- label: ":amd: (MI300) LM Eval Large Models" # TBD - timeout_in_minutes: 65 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_8 - optional: true - num_gpus: 8 - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/model_executor/models/ - - vllm/model_executor/model_loader/ - - vllm/model_executor/layers/quantization/ - - vllm/v1/attention/backends/ - - vllm/v1/attention/selector.py - - vllm/model_executor/layers/layernorm.py - - csrc/ - - vllm/_aiter_ops.py - - vllm/platforms/rocm.py - - tests/evals/ - commands: - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - export PYTORCH_ROCM_ARCH=gfx942 - - pytest -s -v evals/gsm8k/test_gsm8k_correctness.py --config-list-file=configs/models-mi3xx.txt - -- label: ":amd: (MI300) LM Eval Large Models ROCm Harness" # TBD - timeout_in_minutes: 40 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_8 - optional: true - num_gpus: 8 - working_dir: "/vllm-workspace/.buildkite/lm-eval-harness" - source_file_dependencies: - - vllm/model_executor/models/ - - vllm/model_executor/model_loader/ - - vllm/model_executor/layers/quantization/ - - vllm/v1/attention/backends/ - - vllm/v1/attention/selector.py - - vllm/model_executor/layers/layernorm.py - - csrc/ - - vllm/_aiter_ops.py - - vllm/platforms/rocm.py - commands: - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - pytest -s -v test_lm_eval_correctness.py --config-list-file=configs/models-large-rocm.txt --tp-size=8 --timeout=5400 - - label: ":amd: (MI300) LM Eval Large Models FP8" # TBD timeout_in_minutes: 40 mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] @@ -1939,20 +1839,6 @@ steps: commands: - pytest -v -s models/language/pooling_mteb_test -- label: ":amd: (MI300) Language Models (Extended Pooling) Shard %N" - timeout_in_minutes: 95 - parallelism: 4 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_1 - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/models/language/pooling - commands: - - pytest -v -s models/language/pooling -m 'not core_model' --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB - - label: ":amd: (MI300) Language Models (Extended Generation)" # TBD timeout_in_minutes: 105 mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] @@ -1970,21 +1856,6 @@ steps: #---------------------------------------------------- mi300 · models / multimodal ----------------------------------------------------# -- label: ":amd: (MI300) Multimodal Models (Extended Generation 1)" # TBD - timeout_in_minutes: 100 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_1 - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/models/multimodal/generation - - tests/models/multimodal/test_mapping.py - commands: - - pytest -v -s models/multimodal/generation -m 'not core_model' --ignore models/multimodal/generation/test_common.py - - pytest -v -s models/multimodal/test_mapping.py - - label: ":amd: (MI300) Multimodal Models (Extended Generation 2) Shard %N" timeout_in_minutes: 100 parallelism: 4 @@ -2013,19 +1884,6 @@ steps: commands: - pytest -v -s models/multimodal/generation/test_common.py -m 'split(group=1) and not core_model' -- label: ":amd: (MI300) Multimodal Models (PPL)" - timeout_in_minutes: 90 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_1 - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/models/multimodal/generation_ppl_test/ - commands: - - pytest -v -s models/multimodal/generation_ppl_test/ - - label: ":amd: (MI300) Multimodal Models (Standard) 3: llava + qwen2_vl" # TBD timeout_in_minutes: 55 mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] @@ -2057,19 +1915,6 @@ steps: commands: - pytest -v -s models/multimodal/processing/test_tensor_schema.py --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB -- label: ":amd: (MI300) Multimodal Models (Extended Pooling)" # TBD - timeout_in_minutes: 75 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_1 - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/models/multimodal/pooling - commands: - - pytest -v -s models/multimodal/pooling -m 'not core_model' - #----------------------------------------------------- mi300 · models / quantized -----------------------------------------------------# - label: ":amd: (MI300) Quantized Models" # TBD @@ -2581,60 +2426,6 @@ steps: commands: - pytest -v -s v1/e2e/spec_decode/test_mtp_parallel_load.py -- label: ":amd: (MI300) Spec Decode AL DFlash Nightly" - timeout_in_minutes: 120 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_1 - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/v1/spec_decode/ - - vllm/v1/worker/gpu/spec_decode/ - - vllm/model_executor/models/qwen3_dflash.py - - vllm/model_executor/models/qwen3_dflash2.py - - vllm/model_executor/models/qwen3_5.py - - vllm/model_executor/models/laguna_dflash.py - - tests/v1/e2e/spec_decode/ - - vllm/v1/attention/ - - vllm/v1/sample/ - - vllm/model_executor/model_loader/ - - vllm/model_executor/layers/ - - vllm/model_executor/models/qwen2.py - - vllm/model_executor/models/qwen3.py - - vllm/model_executor/models/laguna.py - - vllm/platforms/rocm.py - commands: - - pytest -v -s v1/e2e/spec_decode/acceptance_rates/dflash/ -k "not dflash2" - -- label: ":amd: (MI300) Spec Decode AL DSpark Nightly" - timeout_in_minutes: 90 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_1 - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/v1/spec_decode/ - - vllm/v1/worker/gpu/spec_decode/ - - vllm/model_executor/models/qwen3_dspark.py - - vllm/model_executor/models/gemma4_dspark.py - - tests/v1/e2e/spec_decode/ - - vllm/v1/attention/ - - vllm/v1/sample/ - - vllm/model_executor/model_loader/ - - vllm/model_executor/layers/ - - vllm/model_executor/models/qwen2.py - - vllm/model_executor/models/qwen3.py - - vllm/model_executor/models/qwen3_dflash.py - - vllm/model_executor/models/gemma4.py - - vllm/model_executor/models/gemma4_mm.py - - vllm/model_executor/models/gemma4_unified.py - - vllm/model_executor/models/gemma4_mtp.py - - vllm/platforms/rocm.py - commands: - - pytest -v -s v1/e2e/spec_decode/acceptance_rates/dspark/ - - label: ":amd: (MI300) V1 Spec Decode" # TBD timeout_in_minutes: 50 mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] @@ -2719,39 +2510,6 @@ steps: - pytest -v -s v1/engine --ignore v1/engine/test_preprocess_error_handling.py - pytest -v -s v1/test_tensor_ipc_queue.py -- label: ":amd: (MI300) Speculators Correctness Nightly" - timeout_in_minutes: 40 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_1 - optional: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/config/speculative.py - - vllm/engine/arg_utils.py - - vllm/transformers_utils/config.py - - vllm/transformers_utils/configs/speculators/ - - vllm/v1/spec_decode/ - - vllm/v1/worker/gpu/spec_decode/ - - vllm/v1/worker/gpu_model_runner.py - - vllm/v1/sample/ - - vllm/v1/attention/backends/ - - vllm/v1/attention/selector.py - - vllm/model_executor/model_loader/ - - vllm/model_executor/layers/ - - vllm/model_executor/models/llama_eagle3.py - - vllm/model_executor/models/qwen3.py - - vllm/model_executor/models/qwen3_dflash.py - - vllm/model_executor/models/registry.py - - vllm/_aiter_ops.py - - tests/evals/gsm8k/ - - tests/v1/spec_decode/test_speculators_correctness.py - - vllm/platforms/rocm.py - - vllm/v1/worker/gpu/spec_decode/dflash/ - commands: - - export VLLM_ALLOW_INSECURE_SERIALIZATION=1 - - pytest -v -s v1/spec_decode/test_speculators_correctness.py -m slow_test - - label: ":amd: (MI300) Extract Hidden States Integration" # TBD timeout_in_minutes: 30 mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] @@ -3106,22 +2864,6 @@ steps: - uv pip install --system -r /vllm-workspace/requirements/kv_connectors_rocm.txt - ATTENTION_BACKEND=TRITON_ATTN bash v1/kv_connector/nixl_integration/run_multi_connector_edge_case_test.sh -- label: ":amd: (MI300) V1 E2E Hybrid Chunked Prefill" # TBD - timeout_in_minutes: 35 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_4 - num_gpus: 4 - optional: true - source_file_dependencies: - - vllm/v1/attention/backends/utils.py - - vllm/v1/worker/gpu_model_runner.py - - tests/v1/e2e/test_hybrid_chunked_prefill.py - - vllm/v1/core/ - working_dir: /vllm-workspace/tests - commands: - - pytest -v -s v1/e2e/test_hybrid_chunked_prefill.py - ######################################################################################################################################### # # # MI355 (gfx950) tests # @@ -4604,6 +4346,82 @@ steps: - export VLLM_WORKER_MULTIPROC_METHOD=spawn - pytest -s -v evals/gsm8k/test_gsm8k_correctness.py --config-list-file=configs/models-mi355-nemotron-3-super.txt +- label: ":amd: (MI355) LM Eval Small Models" # TBD + timeout_in_minutes: 45 + mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] + dind: false + agent_pool: mi355_dpx + working_dir: "/vllm-workspace/tests" + source_file_dependencies: + - csrc/ + - vllm/model_executor/layers/quantization + - vllm/model_executor/models/ + - vllm/model_executor/model_loader/ + - vllm/v1/attention/backends/ + - vllm/v1/attention/selector.py + - vllm/_aiter_ops.py + - vllm/platforms/rocm.py + optional: true + commands: + - pytest -s -v evals/gsm8k/test_gsm8k_correctness.py --config-list-file=configs/models-small.txt + +- label: ":amd: (MI355 DPX) LM Eval KV-Offload" # TBD + timeout_in_minutes: 55 + mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] + dind: false + agent_pool: mi355_dpx + optional: true + working_dir: "/vllm-workspace/tests" + source_file_dependencies: + - vllm/distributed/kv_transfer/kv_connector/v1/offloading/ + - vllm/distributed/kv_transfer/kv_connector/v1/simple_cpu_offload_connector.py + - vllm/v1/kv_offload/ + - vllm/v1/simple_kv_offload/ + - tests/evals/gsm8k/test_gsm8k_offloading.py + - vllm/platforms/rocm.py + commands: + - pytest -s -v evals/gsm8k/test_gsm8k_offloading.py -k "nemotron-h-8b or gemma-4-e4b-it" + +- label: ":amd: (MI355) LM Eval KV-Offload Large" # TBD + timeout_in_minutes: 65 + mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] + dind: false + agent_pool: mi355_4 + num_gpus: 4 + optional: true + working_dir: "/vllm-workspace/tests" + source_file_dependencies: + - vllm/distributed/kv_transfer/kv_connector/v1/offloading/ + - vllm/distributed/kv_transfer/kv_connector/v1/simple_cpu_offload_connector.py + - vllm/v1/kv_offload/ + - vllm/v1/simple_kv_offload/ + - tests/evals/gsm8k/test_gsm8k_offloading.py + - vllm/platforms/rocm.py + commands: + - VLLM_ROCM_USE_AITER=1 pytest -s -v evals/gsm8k/test_gsm8k_offloading.py -k "deepseek-v4-flash" + +- label: ":amd: (MI355) LM Eval Large Models ROCm Harness" # TBD + timeout_in_minutes: 40 + mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] + dind: false + agent_pool: mi355_8 + optional: true + num_gpus: 8 + working_dir: "/vllm-workspace/.buildkite/lm-eval-harness" + source_file_dependencies: + - vllm/model_executor/models/ + - vllm/model_executor/model_loader/ + - vllm/model_executor/layers/quantization/ + - vllm/v1/attention/backends/ + - vllm/v1/attention/selector.py + - vllm/model_executor/layers/layernorm.py + - csrc/ + - vllm/_aiter_ops.py + - vllm/platforms/rocm.py + commands: + - export VLLM_WORKER_MULTIPROC_METHOD=spawn + - pytest -s -v test_lm_eval_correctness.py --config-list-file=configs/models-large-rocm.txt --tp-size=8 --timeout=5400 + #--------------------------------------------------------- mi355 · examples ----------------------------------------------------------# - label: ":amd: (MI355 DPX) Examples" # TBD @@ -5421,6 +5239,19 @@ steps: - pytest models/multimodal/generation/test_memory_leak.py -m core_model - cd .. && VLLM_WORKER_MULTIPROC_METHOD=spawn pytest -v -s tests/models/multimodal/generation/test_whisper.py -m core_model +- label: ":amd: (MI355 DPX) Multimodal Models (PPL)" + timeout_in_minutes: 90 + 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/generation_ppl_test/ + commands: + - pytest -v -s models/multimodal/generation_ppl_test/ + #----------------------------------------------------- mi355 · models / quantized -----------------------------------------------------# - label: ":amd: (MI355 DPX) Quantized Models" # TBD @@ -6041,6 +5872,108 @@ steps: - if [ "$$BUILDKITE_PARALLEL_JOB" = "1" ]; then pytest -v -s "v1/e2e/spec_decode/acceptance_rates/mtp_other/test_mtp.py::test_gemma4_mtp_acceptance_lengths[False]"; fi - if [ "$$BUILDKITE_PARALLEL_JOB" = "2" ]; then pytest -v -s "v1/e2e/spec_decode/acceptance_rates/mtp_other/test_mtp.py::test_gemma4_mtp_acceptance_lengths[True]"; fi +- label: ":amd: (MI355 DPX) V1 E2E Hybrid Chunked Prefill" # TBD + timeout_in_minutes: 35 + mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] + dind: false + agent_pool: mi355_dpx + optional: true + source_file_dependencies: + - vllm/v1/attention/backends/utils.py + - vllm/v1/worker/gpu_model_runner.py + - tests/v1/e2e/test_hybrid_chunked_prefill.py + - vllm/v1/core/ + working_dir: /vllm-workspace/tests + commands: + - pytest -v -s v1/e2e/test_hybrid_chunked_prefill.py + +- label: ":amd: (MI355 DPX) Speculators Correctness Nightly" + timeout_in_minutes: 40 + mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] + dind: false + agent_pool: mi355_dpx + optional: true + working_dir: "/vllm-workspace/tests" + source_file_dependencies: + - vllm/config/speculative.py + - vllm/engine/arg_utils.py + - vllm/transformers_utils/config.py + - vllm/transformers_utils/configs/speculators/ + - vllm/v1/spec_decode/ + - vllm/v1/worker/gpu/spec_decode/ + - vllm/v1/worker/gpu_model_runner.py + - vllm/v1/sample/ + - vllm/v1/attention/backends/ + - vllm/v1/attention/selector.py + - vllm/model_executor/model_loader/ + - vllm/model_executor/layers/ + - vllm/model_executor/models/llama_eagle3.py + - vllm/model_executor/models/qwen3.py + - vllm/model_executor/models/qwen3_dflash.py + - vllm/model_executor/models/registry.py + - vllm/_aiter_ops.py + - tests/evals/gsm8k/ + - tests/v1/spec_decode/test_speculators_correctness.py + - vllm/platforms/rocm.py + - vllm/v1/worker/gpu/spec_decode/dflash/ + commands: + - export VLLM_ALLOW_INSECURE_SERIALIZATION=1 + - pytest -v -s v1/spec_decode/test_speculators_correctness.py -m slow_test + +- label: ":amd: (MI355 DPX) Spec Decode AL DFlash Nightly" + timeout_in_minutes: 120 + mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] + dind: false + agent_pool: mi355_dpx + optional: true + working_dir: "/vllm-workspace/tests" + source_file_dependencies: + - vllm/v1/spec_decode/ + - vllm/v1/worker/gpu/spec_decode/ + - vllm/model_executor/models/qwen3_dflash.py + - vllm/model_executor/models/qwen3_dflash2.py + - vllm/model_executor/models/qwen3_5.py + - vllm/model_executor/models/laguna_dflash.py + - tests/v1/e2e/spec_decode/ + - vllm/v1/attention/ + - vllm/v1/sample/ + - vllm/model_executor/model_loader/ + - vllm/model_executor/layers/ + - vllm/model_executor/models/qwen2.py + - vllm/model_executor/models/qwen3.py + - vllm/model_executor/models/laguna.py + - vllm/platforms/rocm.py + commands: + - pytest -v -s v1/e2e/spec_decode/acceptance_rates/dflash/ -k "not dflash2" + +- label: ":amd: (MI355 DPX) Spec Decode AL DSpark Nightly" + timeout_in_minutes: 90 + mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] + dind: false + agent_pool: mi355_dpx + optional: true + working_dir: "/vllm-workspace/tests" + source_file_dependencies: + - vllm/v1/spec_decode/ + - vllm/v1/worker/gpu/spec_decode/ + - vllm/model_executor/models/qwen3_dspark.py + - vllm/model_executor/models/gemma4_dspark.py + - tests/v1/e2e/spec_decode/ + - vllm/v1/attention/ + - vllm/v1/sample/ + - vllm/model_executor/model_loader/ + - vllm/model_executor/layers/ + - vllm/model_executor/models/qwen2.py + - vllm/model_executor/models/qwen3.py + - vllm/model_executor/models/qwen3_dflash.py + - vllm/model_executor/models/gemma4.py + - vllm/model_executor/models/gemma4_mm.py + - vllm/model_executor/models/gemma4_unified.py + - vllm/model_executor/models/gemma4_mtp.py + - vllm/platforms/rocm.py + commands: + - pytest -v -s v1/e2e/spec_decode/acceptance_rates/dspark/ + #----------------------------------------------------------- mi355 · misc ------------------------------------------------------------# - label: ":amd: (MI355 DPX) Regression" # TBD diff --git a/.buildkite/test_areas/engine.yaml b/.buildkite/test_areas/engine.yaml index bddee97a3a..ef4b7951ef 100644 --- a/.buildkite/test_areas/engine.yaml +++ b/.buildkite/test_areas/engine.yaml @@ -238,9 +238,9 @@ steps: - pytest -v -s v1/e2e/test_hybrid_chunked_prefill.py mirror: amd: - label: ":amd: (MI300) V1 E2E Hybrid Chunked Prefill" + label: ":amd: (MI355 DPX) V1 E2E Hybrid Chunked Prefill" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 35 depends_on: - image-build-amd diff --git a/.buildkite/test_areas/lm_eval.yaml b/.buildkite/test_areas/lm_eval.yaml index af0b419de9..cb4e7eaedc 100644 --- a/.buildkite/test_areas/lm_eval.yaml +++ b/.buildkite/test_areas/lm_eval.yaml @@ -15,9 +15,9 @@ steps: - pytest -s -v evals/gsm8k/test_gsm8k_correctness.py --config-list-file=configs/models-small.txt mirror: amd: - label: ":amd: (MI300) LM Eval Small Models" + label: ":amd: (MI355) LM Eval Small Models" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 45 depends_on: - image-build-amd @@ -707,18 +707,17 @@ steps: - export VLLM_WORKER_MULTIPROC_METHOD=spawn - pytest -s -v evals/gsm8k/test_gsm8k_correctness.py --config-list-file=configs/models-mi355-nemotron-3-super.txt -- label: ":amd: (MI300) LM Eval Large Models" +- label: ":amd: (MI355) LM Eval Large Models" key: amd-lm-eval-large-models dind: false - device: mi300_8 + device: mi355_8 optional: true timeout_in_minutes: 65 depends_on: - image-build-amd commands: - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - export PYTORCH_ROCM_ARCH=gfx942 # Limit Quark compilation to save time - - pytest -s -v evals/gsm8k/test_gsm8k_correctness.py --config-list-file=configs/models-mi3xx.txt + - pytest -s -v evals/gsm8k/test_gsm8k_correctness.py --config-list-file=configs/models-gfx950-large.txt source_file_dependencies: - vllm/_aiter_ops.py - vllm/platforms/rocm.py @@ -1298,9 +1297,9 @@ steps: - pytest -s -v evals/gsm8k/test_gsm8k_offloading.py -k "nemotron-h-8b or gemma-4-e4b-it" mirror: amd: - label: ":amd: (MI300) LM Eval KV-Offload" + label: ":amd: (MI355 DPX) LM Eval KV-Offload" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 55 depends_on: - image-build-amd @@ -1360,9 +1359,9 @@ steps: - pytest -s -v evals/gsm8k/test_gsm8k_offloading.py -k "deepseek-v4-flash" mirror: amd: - label: ":amd: (MI300) LM Eval KV-Offload Large" + label: ":amd: (MI355) LM Eval KV-Offload Large" dind: false - device: mi300_4 + device: mi355_4 timeout_in_minutes: 65 depends_on: - image-build-amd @@ -1427,10 +1426,10 @@ steps: # Qwen2-57B has 28 attention heads and cannot run at TP8. - pytest -s -v test_lm_eval_correctness.py --config-list-file=configs/models-large.txt --tp-size=4 -- label: ":amd: (MI300) LM Eval Large Models ROCm Harness" +- label: ":amd: (MI355) LM Eval Large Models ROCm Harness" key: amd-lm-eval-large-models-rocm-harness dind: false - device: mi300_8 + device: mi355_8 optional: true timeout_in_minutes: 40 depends_on: diff --git a/.buildkite/test_areas/models_language.yaml b/.buildkite/test_areas/models_language.yaml index c30a25993c..be9f6927b3 100644 --- a/.buildkite/test_areas/models_language.yaml +++ b/.buildkite/test_areas/models_language.yaml @@ -216,9 +216,9 @@ steps: parallelism: 4 commands: - "pytest -v -s models/language/pooling -m 'not core_model' --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB" - label: ":amd: (MI300) Language Models (Extended Pooling) Shard %N" + label: ":amd: (MI355 DPX) Language Models (Extended Pooling) Shard %N" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 95 depends_on: - image-build-amd diff --git a/.buildkite/test_areas/models_multimodal.yaml b/.buildkite/test_areas/models_multimodal.yaml index d281d16f7f..2cb2079d9a 100644 --- a/.buildkite/test_areas/models_multimodal.yaml +++ b/.buildkite/test_areas/models_multimodal.yaml @@ -163,9 +163,9 @@ steps: - pytest -v -s models/multimodal/test_mapping.py mirror: amd: - label: ":amd: (MI300) Multimodal Models (Extended Generation 1)" + label: ":amd: (MI355 DPX) Multimodal Models (Extended Generation 1)" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 100 depends_on: - image-build-amd @@ -182,9 +182,9 @@ steps: - pytest -v -s models/multimodal/generation_ppl_test/ mirror: amd: - label: ":amd: (MI300) Multimodal Models (PPL)" + label: ":amd: (MI355 DPX) Multimodal Models (PPL)" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 90 depends_on: - image-build-amd @@ -240,9 +240,9 @@ steps: - pytest -v -s models/multimodal/pooling -m 'not core_model' mirror: amd: - label: ":amd: (MI300) Multimodal Models (Extended Pooling)" + label: ":amd: (MI355 DPX) Multimodal Models (Extended Pooling)" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 75 depends_on: - image-build-amd diff --git a/.buildkite/test_areas/spec_decode.yaml b/.buildkite/test_areas/spec_decode.yaml index c2742b4c51..bfc81b7d7f 100644 --- a/.buildkite/test_areas/spec_decode.yaml +++ b/.buildkite/test_areas/spec_decode.yaml @@ -137,9 +137,9 @@ steps: - pytest -v -s v1/e2e/spec_decode/speculators/ mirror: amd: - label: ":amd: (MI300) Spec Decode Speculators" + label: ":amd: (MI355 DPX) Spec Decode Speculators" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 40 depends_on: - image-build-amd @@ -167,9 +167,9 @@ steps: - pytest -v -s v1/e2e/spec_decode/mtp/gemma4/ mirror: amd: - label: ":amd: (MI300) Spec Decode MTP Gemma4" + label: ":amd: (MI355 DPX) Spec Decode MTP Gemma4" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 40 depends_on: - image-build-amd @@ -198,9 +198,9 @@ steps: - pytest -v -s v1/e2e/spec_decode/mtp/mimo/ mirror: amd: - label: ":amd: (MI300) Spec Decode MTP DeepSeek + MiMo" + label: ":amd: (MI355 DPX) Spec Decode MTP DeepSeek + MiMo" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 40 depends_on: - image-build-amd @@ -228,9 +228,9 @@ steps: - pytest -v -s v1/e2e/spec_decode/mtp/qwen3_5/ mirror: amd: - label: ":amd: (MI300) Spec Decode MTP Qwen3.5" + label: ":amd: (MI355 DPX) Spec Decode MTP Qwen3.5" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 40 depends_on: - image-build-amd @@ -387,9 +387,9 @@ steps: - pytest -v -s v1/spec_decode/test_speculators_correctness.py -m slow_test mirror: amd: - label: ":amd: (MI300) Speculators Correctness Nightly" + label: ":amd: (MI355 DPX) Speculators Correctness Nightly" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 40 depends_on: - image-build-amd @@ -452,9 +452,9 @@ steps: - pytest -v -s v1/e2e/spec_decode/acceptance_rates/dflash/ -k "not dflash2" mirror: amd: - label: ":amd: (MI300) Spec Decode AL DFlash Nightly" + label: ":amd: (MI355 DPX) Spec Decode AL DFlash Nightly" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 120 commands: - pytest -v -s v1/e2e/spec_decode/acceptance_rates/dflash/ -k "not dflash2" @@ -531,9 +531,9 @@ steps: - pytest -v -s v1/e2e/spec_decode/acceptance_rates/dspark/ mirror: amd: - label: ":amd: (MI300) Spec Decode AL DSpark Nightly" + label: ":amd: (MI355 DPX) Spec Decode AL DSpark Nightly" dind: false - device: mi300_1 + device: mi355_dpx timeout_in_minutes: 90 depends_on: - image-build-amd diff --git a/tests/evals/gsm8k/configs/DeepSeek-R1-DP_MI325.yaml b/tests/evals/gsm8k/configs/DeepSeek-R1-DP_MI325.yaml deleted file mode 100644 index 0171cb4b19..0000000000 --- a/tests/evals/gsm8k/configs/DeepSeek-R1-DP_MI325.yaml +++ /dev/null @@ -1,12 +0,0 @@ -model_name: "deepseek-ai/DeepSeek-R1" -accuracy_threshold: 0.95 -num_questions: 1319 -num_fewshot: 5 -startup_max_wait_seconds: 1200 -server_args: >- - --enforce-eager - --max-model-len 4096 - --data-parallel-size 8 - --enable-expert-parallel - --attention-backend=TRITON_ATTN - --speculative-config '{"method":"mtp","num_speculative_tokens":3}' diff --git a/tests/evals/gsm8k/configs/DeepSeek-R1-TP_MI325.yaml b/tests/evals/gsm8k/configs/DeepSeek-R1-TP_MI325.yaml deleted file mode 100644 index ef92f574c7..0000000000 --- a/tests/evals/gsm8k/configs/DeepSeek-R1-TP_MI325.yaml +++ /dev/null @@ -1,12 +0,0 @@ -model_name: "deepseek-ai/DeepSeek-R1" -accuracy_threshold: 0.95 -num_questions: 1319 -num_fewshot: 5 -startup_max_wait_seconds: 1200 -server_args: >- - --enforce-eager - --max-model-len 4096 - --tensor-parallel-size 8 - --enable-expert-parallel - --attention-backend=TRITON_ATTN - --speculative-config '{"method":"mtp","num_speculative_tokens":3}' diff --git a/tests/evals/gsm8k/configs/DeepSeek-V3.2-DP_MI325.yaml b/tests/evals/gsm8k/configs/DeepSeek-V3.2-DP_MI325.yaml deleted file mode 100644 index 8d207878d4..0000000000 --- a/tests/evals/gsm8k/configs/DeepSeek-V3.2-DP_MI325.yaml +++ /dev/null @@ -1,12 +0,0 @@ -model_name: "deepseek-ai/DeepSeek-V3.2" -accuracy_threshold: 0.95 -num_questions: 1319 -num_fewshot: 5 -startup_max_wait_seconds: 1200 -server_args: >- - --enforce-eager - --max-model-len 4096 - --data-parallel-size 8 - --enable-expert-parallel - --attention-backend=TRITON_ATTN - --speculative-config '{"method":"mtp","num_speculative_tokens":3}' diff --git a/tests/evals/gsm8k/configs/DeepSeek-V3.2-TP_MI325.yaml b/tests/evals/gsm8k/configs/DeepSeek-V3.2-TP_MI325.yaml deleted file mode 100644 index 46853d3f5e..0000000000 --- a/tests/evals/gsm8k/configs/DeepSeek-V3.2-TP_MI325.yaml +++ /dev/null @@ -1,12 +0,0 @@ -model_name: "deepseek-ai/DeepSeek-V3.2" -accuracy_threshold: 0.95 -num_questions: 1319 -num_fewshot: 5 -startup_max_wait_seconds: 1200 -server_args: >- - --enforce-eager - --max-model-len 4096 - --tensor-parallel-size 8 - --enable-expert-parallel - --attention-backend=TRITON_ATTN - --speculative-config '{"method":"mtp","num_speculative_tokens":3}' diff --git a/tests/evals/gsm8k/configs/Nemotron-3-Super-120B-A12B-BF16_MI355.yaml b/tests/evals/gsm8k/configs/Nemotron-3-Super-120B-A12B-BF16_MI355.yaml index bf67ca1f7d..44fc3e3bb0 100644 --- a/tests/evals/gsm8k/configs/Nemotron-3-Super-120B-A12B-BF16_MI355.yaml +++ b/tests/evals/gsm8k/configs/Nemotron-3-Super-120B-A12B-BF16_MI355.yaml @@ -12,4 +12,5 @@ server_args: >- --enable-expert-parallel --mamba-backend triton --moe-backend triton + --attention-backend ROCM_AITER_FA --speculative-config '{"method":"mtp","num_speculative_tokens":5}' diff --git a/tests/evals/gsm8k/configs/Nemotron-3-Super-120B-A12B-FP8_MI355.yaml b/tests/evals/gsm8k/configs/Nemotron-3-Super-120B-A12B-FP8_MI355.yaml index 9d76e8f3ab..877b4e7e2f 100644 --- a/tests/evals/gsm8k/configs/Nemotron-3-Super-120B-A12B-FP8_MI355.yaml +++ b/tests/evals/gsm8k/configs/Nemotron-3-Super-120B-A12B-FP8_MI355.yaml @@ -12,4 +12,5 @@ server_args: >- --enable-expert-parallel --mamba-backend triton --moe-backend triton + --attention-backend ROCM_AITER_FA --speculative-config '{"method":"mtp","num_speculative_tokens":5}' diff --git a/tests/evals/gsm8k/configs/models-mi3xx.txt b/tests/evals/gsm8k/configs/models-mi3xx.txt deleted file mode 100644 index 70d9be32e3..0000000000 --- a/tests/evals/gsm8k/configs/models-mi3xx.txt +++ /dev/null @@ -1,8 +0,0 @@ -DeepSeek-R1-TP_MI325.yaml -DeepSeek-R1-DP_MI325.yaml -DeepSeek-V3.2-TP_MI325.yaml -DeepSeek-V3.2-DP_MI325.yaml -Qwen3-30B-A3B-NVFP4.yaml -Qwen3-30B-A3B-NVFP4-quark.yaml -Qwen3.5-35B-A3B-MXFP4-AITER-TP2.yaml -Qwen3.5-35B-A3B-MXFP4-EMU-TP2.yaml