ROCm · gshtras · Jan 8, 2025 · Dec 16, 2024 · Dec 16, 2024 · Dec 16, 2024
diff --git a/.buildkite/generate_index.py b/.buildkite/generate_index.py
@@ -0,0 +1,24 @@
+import argparse
+import os
+
+template = """<!DOCTYPE html>
+<html>
+    <body>
+    <h1>Links for vLLM</h1/>
+        <a href="../{wheel_html_escaped}">{wheel}</a><br/>
+    </body>
+</html>
+"""
+
+parser = argparse.ArgumentParser()
+parser.add_argument("--wheel", help="The wheel path.", required=True)
+args = parser.parse_args()
+
+filename = os.path.basename(args.wheel)
+
+with open("index.html", "w") as f:
+    print(f"Generated index.html for {args.wheel}")
+    # cloudfront requires escaping the '+' character
+    f.write(
+        template.format(wheel=filename,
+                        wheel_html_escaped=filename.replace("+", "%2B")))
diff --git a/.buildkite/nightly-benchmarks/benchmark-pipeline.yaml b/.buildkite/nightly-benchmarks/benchmark-pipeline.yaml
@@ -1,5 +1,6 @@
 steps:
   - label: "Wait for container to be ready"
+    key: wait-for-container-image
     agents:
       queue: A100
     plugins:
@@ -10,12 +11,11 @@ steps:
             command:
             - sh .buildkite/nightly-benchmarks/scripts/wait-for-image.sh
 
-  - wait
-
   - label: "A100"
     # skip: "use this flag to conditionally skip the benchmark step, useful for PR testing"
     agents:
       queue: A100
+    depends_on: wait-for-container-image
     plugins:
     - kubernetes:
         podSpec:
@@ -49,6 +49,7 @@ steps:
     # skip: "use this flag to conditionally skip the benchmark step, useful for PR testing"
     agents:
       queue: H200
+    depends_on: wait-for-container-image
     plugins:
     - docker#v5.12.0:
         image: public.ecr.aws/q9t5s3a7/vllm-ci-postmerge-repo:$BUILDKITE_COMMIT
@@ -65,15 +66,15 @@ steps:
         - VLLM_USAGE_SOURCE
         - HF_TOKEN
 
-  - block: "Run H100 Benchmark"
-    key: block-h100
-    depends_on: ~
+  #- block: "Run H100 Benchmark"
+    #key: block-h100
+    #depends_on: ~
 
   - label: "H100"
     # skip: "use this flag to conditionally skip the benchmark step, useful for PR testing"
     agents:
       queue: H100
-    depends_on: block-h100
+    depends_on: wait-for-container-image
     plugins:
     - docker#v5.12.0:
         image: public.ecr.aws/q9t5s3a7/vllm-ci-postmerge-repo:$BUILDKITE_COMMIT

diff --git a/.buildkite/release-pipeline.yaml b/.buildkite/release-pipeline.yaml
@@ -55,3 +55,18 @@ steps:
           password-env: DOCKERHUB_TOKEN
     env:
       DOCKER_BUILDKIT: "1"
+
+  - block: "Build CPU release image"
+    key: block-cpu-release-image-build
+    depends_on: ~
+
+  - label: "Build and publish CPU release image"
+    depends_on: block-cpu-release-image-build
+    agents:
+      queue: cpu_queue_postmerge
+    commands:
+      - "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7"
+      - "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg GIT_REPO_CHECK=1 --tag public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:$RELEASE_VERSION --progress plain -f Dockerfile.cpu ."
+      - "docker push public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:$RELEASE_VERSION"
+    env:
+      DOCKER_BUILDKIT: "1"
diff --git a/.buildkite/run-gh200-test.sh b/.buildkite/run-gh200-test.sh
@@ -0,0 +1,28 @@
+#!/bin/bash
+
+# This script build the GH200 docker image and run the offline inference inside the container.
+# It serves a sanity check for compilation and basic model usage.
+set -ex
+
+# Skip the new torch installation during build since we are using the specified version for arm64 in the Dockerfile
+python3 use_existing_torch.py
+
+# Try building the docker image
+DOCKER_BUILDKIT=1 docker build . \
+  --target vllm-openai \
+  --platform "linux/arm64" \
+  -t gh200-test \
+  --build-arg max_jobs=66 \
+  --build-arg nvcc_threads=2 \
+  --build-arg torch_cuda_arch_list="9.0+PTX" \
+  --build-arg vllm_fa_cmake_gpu_arches="90-real"
+
+# Setup cleanup
+remove_docker_container() { docker rm -f gh200-test || true; }
+trap remove_docker_container EXIT
+remove_docker_container
+
+# Run the image and test offline inference
+docker run --name gh200-test --gpus=all --entrypoint="" gh200-test bash -c '
+    python3 examples/offline_inference.py
+'
diff --git a/.buildkite/test-pipeline.yaml b/.buildkite/test-pipeline.yaml
@@ -106,14 +106,12 @@ steps:
   source_file_dependencies:
   - vllm/
   commands:
-  - pip install -e ./plugins/vllm_add_dummy_model
   - pytest -v -s entrypoints/llm --ignore=entrypoints/llm/test_lazy_outlines.py --ignore=entrypoints/llm/test_generate.py --ignore=entrypoints/llm/test_generate_multiple_loras.py --ignore=entrypoints/llm/test_guided_generate.py
   - pytest -v -s entrypoints/llm/test_lazy_outlines.py # it needs a clean process
   - pytest -v -s entrypoints/llm/test_generate.py # it needs a clean process
   - pytest -v -s entrypoints/llm/test_generate_multiple_loras.py # it needs a clean process
   - pytest -v -s entrypoints/llm/test_guided_generate.py # it needs a clean process
   - pytest -v -s entrypoints/openai --ignore=entrypoints/openai/test_oot_registration.py
-  - pytest -v -s entrypoints/openai/test_oot_registration.py # it needs a clean process
   - pytest -v -s entrypoints/test_chat_utils.py
   - pytest -v -s entrypoints/offline_mode # Needs to avoid interference with other tests
 
@@ -201,7 +199,7 @@ steps:
     - python3 offline_inference_classification.py
     - python3 offline_inference_embedding.py
     - python3 offline_inference_scoring.py
-    - python3 offline_profile.py --model facebook/opt-125m
+    - python3 offline_profile.py --model facebook/opt-125m run_num_steps --num-steps 2
 
 - label: Prefix Caching Test # 9min
   mirror_hardwares: [amd]
@@ -224,8 +222,12 @@ steps:
   mirror_hardwares: [amd]
   source_file_dependencies:
   - vllm/model_executor/layers
+  - vllm/model_executor/guided_decoding
   - tests/test_logits_processor
-  command: pytest -v -s test_logits_processor.py
+  - tests/model_executor/test_guided_processors
+  commands: 
+    - pytest -v -s test_logits_processor.py
+    - pytest -v -s model_executor/test_guided_processors.py
 
 - label: Speculative decoding tests # 30min
   source_file_dependencies:
@@ -329,8 +331,6 @@ steps:
   - vllm/
   - tests/models
   commands:
-    - pip install -e ./plugins/vllm_add_dummy_model
-    - pytest -v -s models/test_oot_registration.py # it needs a clean process
     - pytest -v -s models/test_registry.py
     - pytest -v -s models/test_initialization.py
 
@@ -356,23 +356,25 @@ steps:
     - pytest -v -s models/decoder_only/language -m 'not core_model and not quant_model'
     - pytest -v -s models/embedding/language -m 'not core_model'
 
-- label: Multi-Modal Models Test (Standard) # 28min
+- label: Multi-Modal Models Test (Standard) # 40min
   #mirror_hardwares: [amd]
   source_file_dependencies:
   - vllm/
   - tests/models/decoder_only/audio_language
   - tests/models/decoder_only/vision_language
   - tests/models/embedding/vision_language
+  - tests/models/encoder_decoder/audio_language
   - tests/models/encoder_decoder/vision_language
   commands:
     - pip install git+https://github.com/TIGER-AI-Lab/Mantis.git
     - pytest -v -s models/decoder_only/audio_language -m 'core_model or quant_model'
     - pytest -v -s --ignore models/decoder_only/vision_language/test_phi3v.py models/decoder_only/vision_language -m 'core_model or quant_model'
     - pytest -v -s models/embedding/vision_language -m core_model
+    - pytest -v -s models/encoder_decoder/audio_language -m core_model
     - pytest -v -s models/encoder_decoder/language -m core_model
     - pytest -v -s models/encoder_decoder/vision_language -m core_model
 
-- label: Multi-Modal Models Test (Extended) 1 # 1h16m
+- label: Multi-Modal Models Test (Extended) 1 # 48m
   optional: true
   source_file_dependencies:
   - vllm/
@@ -465,11 +467,28 @@ steps:
   - pytest models/encoder_decoder/vision_language/test_broadcast.py -v -s -m 'distributed(num_gpus=2)'
   - pytest models/decoder_only/vision_language/test_models.py -v -s -m 'distributed(num_gpus=2)'
   - pytest -v -s spec_decode/e2e/test_integration_dist_tp2.py
-  - pip install -e ./plugins/vllm_add_dummy_model
-  - pytest -v -s distributed/test_distributed_oot.py
   - CUDA_VISIBLE_DEVICES=0,1 pytest -v -s test_sharded_state_loader.py
   - CUDA_VISIBLE_DEVICES=0,1 pytest -v -s kv_transfer/disagg_test.py
 
+- label: Plugin Tests (2 GPUs) # 40min
+  working_dir: "/vllm-workspace/tests"
+  num_gpus: 2
+  fast_check: true
+  source_file_dependencies:
+  - vllm/plugins/
+  - tests/plugins/
+  commands:
+  # begin platform plugin tests, all the code in-between runs on dummy platform
+  - pip install -e ./plugins/vllm_add_dummy_platform
+  - pytest -v -s plugins_tests/test_platform_plugins.py
+  - pip uninstall vllm_add_dummy_platform -y
+  # end platform plugin tests
+  # other tests continue here:
+  - pip install -e ./plugins/vllm_add_dummy_model
+  - pytest -v -s distributed/test_distributed_oot.py
+  - pytest -v -s entrypoints/openai/test_oot_registration.py # it needs a clean process
+  - pytest -v -s models/test_oot_registration.py # it needs a clean process
+
 - label: Multi-step Tests (4 GPUs) # 36min
   working_dir: "/vllm-workspace/tests"
   num_gpus: 4

diff --git a/.buildkite/upload-wheels.sh b/.buildkite/upload-wheels.sh
@@ -23,6 +23,8 @@ wheel="$new_wheel"
 version=$(unzip -p "$wheel" '**/METADATA' | grep '^Version: ' | cut -d' ' -f2)
 echo "Version: $version"
 
+normal_wheel="$wheel" # Save the original wheel filename
+
 # If the version contains "dev", rename it to v1.0.0.dev for consistency
 if [[ $version == *dev* ]]; then
     suffix="${version##*.}"
@@ -32,12 +34,38 @@ if [[ $version == *dev* ]]; then
         new_version="1.0.0.dev"
     fi
     new_wheel="${wheel/$version/$new_version}"
-    mv -- "$wheel" "$new_wheel"
+    # use cp to keep both files in the artifacts directory
+    cp -- "$wheel" "$new_wheel"
     wheel="$new_wheel"
     version="$new_version"
 fi
 
 # Upload the wheel to S3
+python3 .buildkite/generate_index.py --wheel "$normal_wheel"
+
+# generate index for this commit
 aws s3 cp "$wheel" "s3://vllm-wheels/$BUILDKITE_COMMIT/"
+aws s3 cp "$normal_wheel" "s3://vllm-wheels/$BUILDKITE_COMMIT/"
+
+if [[ $normal_wheel == *"cu118"* ]]; then
+    # if $normal_wheel matches cu118, do not upload the index.html
+    echo "Skipping index files for cu118 wheels"
+else
+    # only upload index.html for cu12 wheels (default wheels)
+    aws s3 cp index.html "s3://vllm-wheels/$BUILDKITE_COMMIT/vllm/index.html"
+    aws s3 cp "s3://vllm-wheels/nightly/index.html" "s3://vllm-wheels/$BUILDKITE_COMMIT/index.html"
+fi
+
+# generate index for nightly
 aws s3 cp "$wheel" "s3://vllm-wheels/nightly/"
+aws s3 cp "$normal_wheel" "s3://vllm-wheels/nightly/"
+
+if [[ $normal_wheel == *"cu118"* ]]; then
+    # if $normal_wheel matches cu118, do not upload the index.html
+    echo "Skipping index files for cu118 wheels"
+else
+    # only upload index.html for cu12 wheels (default wheels)
+    aws s3 cp index.html "s3://vllm-wheels/nightly/vllm/index.html"
+fi
+
 aws s3 cp "$wheel" "s3://vllm-wheels/$version/"
diff --git a/.github/ISSUE_TEMPLATE/400-bug report.yml → .github/ISSUE_TEMPLATE/400-bug-report.yml b/.github/ISSUE_TEMPLATE/400-bug report.yml → .github/ISSUE_TEMPLATE/400-bug-report.yml
diff --git a/...ub/ISSUE_TEMPLATE/500-feature request.yml → ...ub/ISSUE_TEMPLATE/500-feature-request.yml b/...ub/ISSUE_TEMPLATE/500-feature request.yml → ...ub/ISSUE_TEMPLATE/500-feature-request.yml
diff --git a/.github/ISSUE_TEMPLATE/600-new model.yml → .github/ISSUE_TEMPLATE/600-new-model.yml b/.github/ISSUE_TEMPLATE/600-new model.yml → .github/ISSUE_TEMPLATE/600-new-model.yml
@@ -9,7 +9,7 @@ body:
     value: >
       #### Before submitting an issue, please make sure the issue hasn't been already addressed by searching through [the existing and past issues](https://github.com/vllm-project/vllm/issues?q=is%3Aissue+sort%3Acreated-desc+).
 
-      #### We also highly recommend you read https://docs.vllm.ai/en/latest/models/adding_model.html first to understand how to add a new model.
+      #### We also highly recommend you read https://docs.vllm.ai/en/latest/contributing/model/adding_model.html first to understand how to add a new model.
 - type: textarea
   attributes:
     label: The model to consider.

diff --git a/...E_TEMPLATE/700-performance discussion.yml → ...E_TEMPLATE/700-performance-discussion.yml b/...E_TEMPLATE/700-performance discussion.yml → ...E_TEMPLATE/700-performance-discussion.yml
diff --git a/...ub/ISSUE_TEMPLATE/800-misc discussion.yml → ...ub/ISSUE_TEMPLATE/800-misc-discussion.yml b/...ub/ISSUE_TEMPLATE/800-misc discussion.yml → ...ub/ISSUE_TEMPLATE/800-misc-discussion.yml
diff --git a/.gitignore b/.gitignore
@@ -81,6 +81,8 @@ instance/
 docs/_build/
 docs/source/getting_started/examples/*.rst
 !**/*.template.rst
+docs/source/getting_started/examples/*.md
+!**/*.template.md
 
 # PyBuilder
 .pybuilder/

diff --git a/CMakeLists.txt b/CMakeLists.txt
@@ -240,7 +240,7 @@ if(VLLM_GPU_LANG STREQUAL "CUDA")
   SET(CUTLASS_ENABLE_HEADERS_ONLY ON CACHE BOOL "Enable only the header library")
 
   # Set CUTLASS_REVISION manually -- its revision detection doesn't work in this case.
-  set(CUTLASS_REVISION "v3.5.1" CACHE STRING "CUTLASS revision to use")
+  set(CUTLASS_REVISION "v3.6.0" CACHE STRING "CUTLASS revision to use")
 
   # Use the specified CUTLASS source directory for compilation if VLLM_CUTLASS_SRC_DIR is provided
   if (DEFINED ENV{VLLM_CUTLASS_SRC_DIR})
@@ -257,7 +257,7 @@ if(VLLM_GPU_LANG STREQUAL "CUDA")
     FetchContent_Declare(
         cutlass
         GIT_REPOSITORY https://github.com/nvidia/cutlass.git
-        GIT_TAG v3.5.1
+        GIT_TAG v3.6.0
         GIT_PROGRESS TRUE
 
         # Speed up CUTLASS download by retrieving only the specified GIT_TAG instead of the history.
@@ -275,7 +275,10 @@ if(VLLM_GPU_LANG STREQUAL "CUDA")
     "csrc/quantization/awq/gemm_kernels.cu"
     "csrc/custom_all_reduce.cu"
     "csrc/permute_cols.cu"
-    "csrc/quantization/cutlass_w8a8/scaled_mm_entry.cu")
+    "csrc/quantization/cutlass_w8a8/scaled_mm_entry.cu"
+    "csrc/sparse/cutlass/sparse_scaled_mm_entry.cu"
+    "csrc/sparse/cutlass/sparse_compressor_entry.cu"
+    "csrc/cutlass_extensions/common.cpp")
 
   set_gencode_flags_for_srcs(
     SRCS "${VLLM_EXT_SRC}"
@@ -304,7 +307,6 @@ if(VLLM_GPU_LANG STREQUAL "CUDA")
                    " in CUDA target architectures")
   endif()
 
-  #
   # The cutlass_scaled_mm kernels for Hopper (c3x, i.e. CUTLASS 3.x) require
   # CUDA 12.0 or later (and only work on Hopper, 9.0/9.0a for now).
   cuda_archs_loose_intersection(SCALED_MM_3X_ARCHS "9.0;9.0a" "${CUDA_ARCHS}")
@@ -357,6 +359,31 @@ if(VLLM_GPU_LANG STREQUAL "CUDA")
     endif()
   endif()
 
+  #
+  # 2:4 Sparse Kernels
+
+  # The 2:4 sparse kernels cutlass_scaled_sparse_mm and cutlass_compressor
+  # require CUDA 12.2 or later (and only work on Hopper, 9.0/9.0a for now).
+  if(${CMAKE_CUDA_COMPILER_VERSION} VERSION_GREATER 12.2 AND SCALED_MM_3X_ARCHS)
+    set(SRCS "csrc/sparse/cutlass/sparse_compressor_c3x.cu"
+             "csrc/sparse/cutlass/sparse_scaled_mm_c3x.cu")
+    set_gencode_flags_for_srcs(
+      SRCS "${SRCS}"
+      CUDA_ARCHS "${SCALED_MM_3X_ARCHS}")
+    list(APPEND VLLM_EXT_SRC "${SRCS}")
+    list(APPEND VLLM_GPU_FLAGS "-DENABLE_SPARSE_SCALED_MM_C3X=1")
+    message(STATUS "Building sparse_scaled_mm_c3x for archs: ${SCALED_MM_3X_ARCHS}")
+  else()
+    if (NOT ${CMAKE_CUDA_COMPILER_VERSION} VERSION_GREATER 12.2 AND SCALED_MM_3X_ARCHS)
+      message(STATUS "Not building sparse_scaled_mm_c3x kernels as CUDA Compiler version is "
+                     "not >= 12.2, we recommend upgrading to CUDA 12.2 or later "
+                     "if you intend on running FP8 sparse quantized models on Hopper.")
+    else()
+      message(STATUS "Not building sparse_scaled_mm_c3x as no compatible archs found "
+                     "in CUDA target architectures")
+    endif()
+  endif()
+
 
   #
   # Machete kernels
@@ -443,7 +470,7 @@ define_gpu_extension_target(
   SOURCES ${VLLM_EXT_SRC}
   COMPILE_FLAGS ${VLLM_GPU_FLAGS}
   ARCHITECTURES ${VLLM_GPU_ARCHES}
-  INCLUDE_DIRECTORIES ${CUTLASS_INCLUDE_DIR}
+  INCLUDE_DIRECTORIES ${CUTLASS_INCLUDE_DIR};${CUTLASS_TOOLS_UTIL_INCLUDE_DIR}
   USE_SABI 3
   WITH_SOABI)
 
@@ -583,7 +610,7 @@ else()
   FetchContent_Declare(
           vllm-flash-attn
           GIT_REPOSITORY https://github.com/vllm-project/flash-attention.git
-          GIT_TAG 04325b6798bcc326c86fb35af62d05a9c8c8eceb
+          GIT_TAG 96266b1111111f3d11aabefaf3bacbab6a89d03c
           GIT_PROGRESS TRUE
           # Don't share the vllm-flash-attn build between build types
           BINARY_DIR ${CMAKE_BINARY_DIR}/vllm-flash-attn