Updating broken link (#5258 )

Update PyTorch training benchmark docker doc to 25.7 (#5255 ) (#5260 )
* Update PyTorch training benchmark docker doc to 25.7 * update .wordlist.txt * update conf.py * update data sheet * fix sphinx warnings
2026-01-10 23:28:03 -05:00 · 2025-09-05 16:02:13 -04:00 · 2025-09-05 12:14:09 -04:00 · 2025-08-26 11:42:57 -04:00 · 2025-08-26 11:35:30 -04:00 · 2025-08-22 11:56:27 -04:00
56 changed files with 1510 additions and 1154 deletions
--- a/.azuredevops/components/HIP.yml
+++ b/.azuredevops/components/HIP.yml
@@ -1,29 +1,10 @@
 parameters:
- name: componentName
-  type: string
-  default: hip_clr_combined
 - name: checkoutRepo
  type: string
  default: 'self'
 - name: checkoutRef
  type: string
  default: ''
-# monorepo related parameters
- name: sparseCheckoutDir
-  type: string
-  default: ''
- name: triggerDownstreamJobs
-  type: boolean
-  default: false
- name: downstreamAggregateNames
-  type: string
-  default: ''
- name: buildDependsOn
-  type: object
-  default: null
- name: unifiedBuild
-  type: boolean
-  default: false
 # set to true if doing full build of ROCm stack
 # and dependencies are pulled from same pipeline
 - name: aggregatePipeline
@@ -54,24 +35,93 @@ parameters:
  type: object
  default:
    - llvm-project
-    - ROCR-Runtime
+
+# hip and clr are tightly-coupled
+# run this same template for both repos
+# any changes for clr should just trigger HIP pipeline
+# similarly for hipother repo, for Nvidia backend

 - name: jobMatrix
  type: object
  default:
    buildJobs:
-      - { os: ubuntu2204, packageManager: apt, platform: amd }
-      - { os: ubuntu2204, packageManager: apt, platform: nvidia }
-      - { os: almalinux8, packageManager: dnf, platform: amd }
-      - { os: almalinux8, packageManager: dnf, platform: nvidia }
+      - { os: ubuntu2204, packageManager: apt }
+      - { os: almalinux8, packageManager: dnf }

+# HIP with AMD backend
 jobs:
 - ${{ each job in parameters.jobMatrix.buildJobs }}:
-  - job: ${{ parameters.componentName }}_${{ job.os }}_${{ job.platform }}
-    ${{ if parameters.buildDependsOn }}:
-      dependsOn:
-        - ${{ each build in parameters.buildDependsOn }}:
-          - ${{ build }}_${{ job.os }}
+  - job: hip_clr_combined_${{ job.os }}_amd
+    pool:
+      vmImage: 'ubuntu-22.04'
+    ${{ if eq(job.os, 'almalinux8') }}:
+      container:
+        image: rocmexternalcicd.azurecr.io/manylinux228:latest
+        endpoint: ContainerService3
+    variables:
+    - group: common
+    - template: /.azuredevops/variables-global.yml
+    workspace:
+      clean: all
+    steps:
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-other.yml
+      parameters:
+        aptPackages: ${{ parameters.aptPackages }}
+        pipModules: ${{ parameters.pipModules }}
+        packageManager: ${{ job.packageManager }}
+  # checkout triggering repo (either HIP or clr)
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
+      parameters:
+        checkoutRepo: ${{ parameters.checkoutRepo }}
+  # if this is triggered by HIP repo, matching repo is clr
+  # if this is triggered by clr repo, matching repo is HIP
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
+      parameters:
+        checkoutRepo: matching_repo
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
+      parameters:
+        checkoutRepo: hipother_repo
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
+      parameters:
+        checkoutRef: ${{ parameters.checkoutRef }}
+        dependencyList: ${{ parameters.rocmDependenciesAMD }}
+        aggregatePipeline: ${{ parameters.aggregatePipeline }}
+        os: ${{ job.os }}
+  # compile clr
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/build-cmake.yml
+      parameters:
+        componentName: clr
+        cmakeBuildDir: '$(Build.SourcesDirectory)/clr/build'
+        cmakeSourceDir: '$(Build.SourcesDirectory)/clr'
+        os: ${{ job.os }}
+        useAmdclang: false
+        extraBuildFlags: >-
+          -DHIP_COMMON_DIR=$(Build.SourcesDirectory)/HIP
+          -DHIP_PLATFORM=amd
+          -DCMAKE_PREFIX_PATH=$(Agent.BuildDirectory)/rocm
+          -DROCM_PATH=$(Agent.BuildDirectory)/rocm
+          -DHIPCC_BIN_DIR=$(Agent.BuildDirectory)/rocm/bin
+          -DCLR_BUILD_HIP=ON
+          -DCLR_BUILD_OCL=ON
+          -GNinja
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/manifest.yml
+      parameters:
+        artifactName: amd
+        os: ${{ job.os }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-upload.yml
+      parameters:
+        artifactName: amd
+        os: ${{ job.os }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-links.yml
+    # - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/docker-container.yml
+    #   parameters:
+    #     aptPackages: ${{ parameters.aptPackages }}
+    #     pipModules: ${{ parameters.pipModules }}
+    #     environment: amd
+
+# HIP with Nvidia backend
+- ${{ each job in parameters.jobMatrix.buildJobs }}:
+  - job: hip_clr_combined_${{ job.os }}_nvidia
    pool:
      vmImage: 'ubuntu-22.04'
    ${{ if eq(job.os, 'almalinux8') }}:
@@ -90,45 +140,49 @@ jobs:
        pipModules: ${{ parameters.pipModules }}
        packageManager: ${{ job.packageManager }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
-    # full checkout of rocm-systems superrepo, we need clr, hip, and hipother
+  # checkout triggering repo (either HIP or clr)
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
      parameters:
        checkoutRepo: ${{ parameters.checkoutRepo }}
-        # sparseCheckoutDir: ${{ parameters.sparseCheckoutDir }}
+  # if this is triggered by HIP repo, matching repo is clr
+  # if this is triggered by clr repo, matching repo is HIP
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
+      parameters:
+        checkoutRepo: matching_repo
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
+      parameters:
+        checkoutRepo: hipother_repo
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
      parameters:
        checkoutRef: ${{ parameters.checkoutRef }}
+        dependencyList: ${{ parameters.rocmDependenciesNvidia }}
        aggregatePipeline: ${{ parameters.aggregatePipeline }}
        os: ${{ job.os }}
-        ${{ if eq(job.platform, 'amd') }}:
-          dependencyList: ${{ parameters.rocmDependenciesAMD }}
-        ${{ elseif eq(job.platform, 'nvidia') }}:
-          dependencyList: ${{ parameters.rocmDependenciesNvidia }}
-        ${{ if parameters.triggerDownstreamJobs }}:
-          downstreamAggregateNames: ${{ parameters.downstreamAggregateNames }}
+    - script: 'ls -1R $(Agent.BuildDirectory)/rocm'
+      displayName: 'Artifact listing'
+  # compile clr
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/build-cmake.yml
      parameters:
        componentName: clr
-        cmakeBuildDir: $(Agent.BuildDirectory)/s/projects/clr/build
-        cmakeSourceDir: $(Agent.BuildDirectory)/s/projects/clr
+        cmakeBuildDir: '$(Build.SourcesDirectory)/clr/build'
+        cmakeSourceDir: '$(Build.SourcesDirectory)/clr'
        os: ${{ job.os }}
        useAmdclang: false
        extraBuildFlags: >-
-          -DCMAKE_PREFIX_PATH=$(Agent.BuildDirectory)/rocm
-          -DROCM_PATH=$(Agent.BuildDirectory)/rocm
+          -DHIP_COMMON_DIR=$(Build.SourcesDirectory)/HIP
+          -DHIP_PLATFORM=nvidia
          -DHIPCC_BIN_DIR=$(Agent.BuildDirectory)/rocm/bin
-          -DHIP_COMMON_DIR=$(Agent.BuildDirectory)/s/projects/hip
-          -DHIPNV_DIR=$(Agent.BuildDirectory)/s/projects/hipother/hipnv
-          -DHIP_PLATFORM=${{ job.platform }}
          -DCLR_BUILD_HIP=ON
-          -DCLR_BUILD_OCL=ON
+          -DCLR_BUILD_OCL=OFF
+          -DHIPNV_DIR=$(Build.SourcesDirectory)/hipother/hipnv
          -GNinja
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/manifest.yml
-      parameters:
-        artifactName: ${{ job.platform }}
-        os: ${{ job.os }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-upload.yml
      parameters:
-        artifactName: ${{ job.platform }}
+        artifactName: nvidia
        os: ${{ job.os }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-links.yml
+    # - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/docker-container.yml
+    #   parameters:
+    #     aptPackages: ${{ parameters.aptPackages }}
+    #     pipModules: ${{ parameters.pipModules }}
+    #     environment: nvidia
--- a/.azuredevops/components/MIOpen.yml
+++ b/.azuredevops/components/MIOpen.yml
@@ -123,7 +123,7 @@ jobs:
    - template: /.azuredevops/variables-global.yml
    - name: ROCM_PATH
      value: $(Agent.BuildDirectory)/rocm
-    pool: ${{ variables.MEDIUM_BUILD_POOL }}
+    pool: ${{ variables.HIGH_BUILD_POOL }}
    workspace:
      clean: all
    steps:
@@ -131,7 +131,6 @@ jobs:
      parameters:
        aptPackages: ${{ parameters.aptPackages }}
        pipModules: ${{ parameters.pipModules }}
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-cmake-latest.yml
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
      parameters:
@@ -150,7 +149,6 @@ jobs:
          downstreamAggregateNames: ${{ parameters.downstreamAggregateNames }}
    - task: Bash@3
      displayName: Build and install other dependencies
-      retryCountOnTaskFailure: 3
      inputs:
        targetType: inline
        workingDirectory: $(Agent.BuildDirectory)/s
@@ -212,7 +210,6 @@ jobs:
      parameters:
        aptPackages: ${{ parameters.aptPackages }}
        pipModules: ${{ parameters.pipModules }}
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-cmake-latest.yml
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
      parameters:
@@ -231,7 +228,6 @@ jobs:
          downstreamAggregateNames: ${{ parameters.downstreamAggregateNames }}
    - task: Bash@3
      displayName: Build and install other dependencies
-      retryCountOnTaskFailure: 3
      inputs:
        targetType: inline
        workingDirectory: $(Agent.BuildDirectory)/s
--- a/.azuredevops/components/Tensile.yml
+++ b/.azuredevops/components/Tensile.yml
@@ -171,7 +171,6 @@ jobs:
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
    - task: DownloadPipelineArtifact@2
      displayName: 'Download Pipeline Wheel Files'
-      retryCountOnTaskFailure: 3
      inputs:
        itemPattern: '**/*${{ job.os }}*.whl'
        targetPath: $(Agent.BuildDirectory)
--- a/.azuredevops/components/hipBLASLt.yml
+++ b/.azuredevops/components/hipBLASLt.yml
@@ -35,13 +35,9 @@ parameters:
    - ccache
    - gfortran
    - git
-    - libboost-filesystem-dev
-    - libboost-program-options-dev
    - libdrm-dev
-    - liblapack-dev
    - libmsgpack-dev
    - libnuma-dev
-    - libopenblas-dev
    - ninja-build
    - python3-pip
    - python3-venv
@@ -50,12 +46,6 @@ parameters:
  default:
    - joblib
    - "packaging>=22.0"
-    - pyyaml
-    - msgpack
-    - simplejson
-    - ujson
-    - orjson
-    - yappi
    - --upgrade
 - name: rocmDependencies
  type: object
@@ -91,12 +81,12 @@ parameters:
      - { pool: rocm-ci_medium_build_pool, os: ubuntu2204, packageManager: apt, target: gfx90a }
      - { pool: rocm-ci_medium_build_pool, os: ubuntu2204, packageManager: apt, target: gfx1201 }
      - { pool: rocm-ci_medium_build_pool, os: ubuntu2204, packageManager: apt, target: gfx1100 }
-      #- { pool: rocm-ci_medium_build_pool, os: ubuntu2204, packageManager: apt, target: gfx1030 }
+      - { pool: rocm-ci_medium_build_pool, os: ubuntu2204, packageManager: apt, target: gfx1030 }
      - { pool: rocm-ci_ultra_build_pool, os: almalinux8, packageManager: dnf, target: gfx942 }
      - { pool: rocm-ci_medium_build_pool, os: almalinux8, packageManager: dnf, target: gfx90a }
      - { pool: rocm-ci_medium_build_pool, os: almalinux8, packageManager: dnf, target: gfx1201 }
      - { pool: rocm-ci_medium_build_pool, os: almalinux8, packageManager: dnf, target: gfx1100 }
-      #- { pool: rocm-ci_medium_build_pool, os: almalinux8, packageManager: dnf, target: gfx1030 }
+      - { pool: rocm-ci_medium_build_pool, os: almalinux8, packageManager: dnf, target: gfx1030 }
    testJobs:
      - { os: ubuntu2204, packageManager: apt, target: gfx942 }
      - { os: ubuntu2204, packageManager: apt, target: gfx90a }
@@ -179,7 +169,7 @@ jobs:
          cd $(Agent.BuildDirectory)/temp-deps
          # position-independent LAPACK is required for almalinux8 builds
          cmake -DBUILD_GTEST=OFF -DBUILD_LAPACK=ON -DCMAKE_POSITION_INDEPENDENT_CODE=ON $(Agent.BuildDirectory)/s/deps
-          make -j
+          make
          sudo make install
    - script: |
        mkdir -p $(CCACHE_DIR)
@@ -205,11 +195,7 @@ jobs:
          -DCMAKE_CXX_COMPILER_LAUNCHER=ccache
          -DCMAKE_C_COMPILER_LAUNCHER=ccache
          -DAMDGPU_TARGETS=${{ job.target }}
-          -DGPU_TARGETS=${{ job.target }}
          -DBUILD_CLIENTS_TESTS=ON
-          -DHIPBLASLT_ENABLE_ROCROLLER=ON
-          -DHIPBLASLT_ENABLE_FETCH=ON
-          -DHIPBLASLT_ENABLE_BLIS=OFF
          -GNinja
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/manifest.yml
      parameters:
--- a/.azuredevops/components/hipSPARSE.yml
+++ b/.azuredevops/components/hipSPARSE.yml
@@ -69,7 +69,7 @@ parameters:
      - { os: ubuntu2204, packageManager: apt, target: gfx942 }
      - { os: ubuntu2204, packageManager: apt, target: gfx90a }
      - { os: ubuntu2204, packageManager: apt, target: gfx1201 }
-      #- { os: ubuntu2204, packageManager: apt, target: gfx1030 }
+      - { os: ubuntu2204, packageManager: apt, target: gfx1030 }
      - { os: ubuntu2204, packageManager: apt, target: gfx1100 }
    testJobs:
      - { os: ubuntu2204, packageManager: apt, target: gfx942 }
--- a/.azuredevops/components/hipSPARSELt.yml
+++ b/.azuredevops/components/hipSPARSELt.yml
@@ -113,8 +113,7 @@ jobs:
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
      parameters:
        checkoutRepo: ${{ parameters.checkoutRepo }}
-        # ignore sparse checkout for monorepo case, we want access to hipblaslt directory
-        # sparseCheckoutDir: ${{ parameters.sparseCheckoutDir }}
+        sparseCheckoutDir: ${{ parameters.sparseCheckoutDir }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
      parameters:
        checkoutRef: ${{ parameters.checkoutRef }}
@@ -131,10 +130,7 @@ jobs:
      displayName: Create temp folder for external dependencies
  # hipSPARSELt already has a CMake script for external deps, so we can just run that
  # https://github.com/ROCm/hipSPARSELt/blob/develop/deps/CMakeLists.txt
-    - ${{ if ne(parameters.sparseCheckoutDir, '') }}:
-        script: cmake $(Pipeline.Workspace)/s/projects/hipsparselt/deps
-      ${{ else }}:
-        script: cmake $(Pipeline.Workspace)/s/deps
+    - script: cmake $(Pipeline.Workspace)/s/deps
      displayName: Configure hipSPARSELt external dependencies
      workingDirectory: $(Pipeline.Workspace)/deps
    - script: make
@@ -158,11 +154,7 @@ jobs:
          -DCMAKE_PREFIX_PATH="$(Agent.BuildDirectory)/rocm"
          -DROCM_PATH=$(Agent.BuildDirectory)/rocm
          -DBUILD_CLIENTS_TESTS=ON
-          -DBUILD_USE_LOCAL_TENSILE=OFF
          -GNinja
-        ${{ if ne(parameters.sparseCheckoutDir, '') }}:
-          cmakeSourceDir: $(Build.SourcesDirectory)/projects/hipsparselt
-          cmakeBuildDir: $(Build.SourcesDirectory)/projects/hipsparselt
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/manifest.yml
      parameters:
        componentName: ${{ parameters.componentName }}
--- a/.azuredevops/components/llvm-project.yml
+++ b/.azuredevops/components/llvm-project.yml
@@ -30,7 +30,7 @@ parameters:
  default:
    buildJobs:
      - { os: ubuntu2204, packageManager: apt }
-      # - { os: ubuntu2404, packageManager: apt }
+      - { os: ubuntu2404, packageManager: apt }
      - { os: almalinux8, packageManager: dnf }

 jobs:
--- a/.azuredevops/components/rccl.yml
+++ b/.azuredevops/components/rccl.yml
@@ -76,7 +76,7 @@ jobs:
    - template: /.azuredevops/variables-global.yml
    - name: HIP_ROCCLR_HOME
      value: $(Build.BinariesDirectory)/rocm
-    pool: ${{ variables.MEDIUM_BUILD_POOL }}
+    pool: ${{ variables.HIGH_BUILD_POOL }}
    workspace:
      clean: all
    steps:
--- a/.azuredevops/components/rocBLAS.yml
+++ b/.azuredevops/components/rocBLAS.yml
@@ -84,12 +84,12 @@ parameters:
      - { os: ubuntu2204, packageManager: apt, target: gfx90a }
      - { os: ubuntu2204, packageManager: apt, target: gfx1201 }
      - { os: ubuntu2204, packageManager: apt, target: gfx1100 }
-      #- { os: ubuntu2204, packageManager: apt, target: gfx1030 }
+      - { os: ubuntu2204, packageManager: apt, target: gfx1030 }
      - { os: almalinux8, packageManager: dnf, target: gfx942 }
      - { os: almalinux8, packageManager: dnf, target: gfx90a }
      - { os: almalinux8, packageManager: dnf, target: gfx1201 }
      - { os: almalinux8, packageManager: dnf, target: gfx1100 }
-      #- { os: almalinux8, packageManager: dnf, target: gfx1030 }
+      - { os: almalinux8, packageManager: dnf, target: gfx1030 }
    testJobs:
      - { os: ubuntu2204, packageManager: apt, target: gfx942 }
      - { os: ubuntu2204, packageManager: apt, target: gfx90a }
@@ -115,13 +115,6 @@ parameters:
 #        buildDependsOn:
 #          - rocBLAS_build
 #          - rocPRIM_build
-    # temporary rocblas->hipblas downstream path while the SOLVERs are disabled
-    - hipBLAS:
-      name: hipBLAS
-      sparseCheckoutDir: projects/hipblas
-      skipUnifiedBuild: 'false'
-      buildDependsOn:
-        - rocBLAS_build

 jobs:
 - ${{ each job in parameters.jobMatrix.buildJobs }}:
--- a/.azuredevops/components/rocPyDecode.yml
+++ b/.azuredevops/components/rocPyDecode.yml
@@ -190,7 +190,6 @@ jobs:
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
    - task: DownloadPipelineArtifact@2
      displayName: 'Download Pipeline Wheel Files'
-      retryCountOnTaskFailure: 3
      inputs:
        itemPattern: '**/*.whl'
        targetPath: $(Agent.BuildDirectory)
--- a/.azuredevops/components/rocSOLVER.yml
+++ b/.azuredevops/components/rocSOLVER.yml
@@ -74,12 +74,12 @@ parameters:
      - { os: ubuntu2204, packageManager: apt, target: gfx90a }
      - { os: ubuntu2204, packageManager: apt, target: gfx1201 }
      - { os: ubuntu2204, packageManager: apt, target: gfx1100 }
-      #- { os: ubuntu2204, packageManager: apt, target: gfx1030 }
+      - { os: ubuntu2204, packageManager: apt, target: gfx1030 }
      - { os: almalinux8, packageManager: dnf, target: gfx942 }
      - { os: almalinux8, packageManager: dnf, target: gfx90a }
      - { os: almalinux8, packageManager: dnf, target: gfx1201 }
      - { os: almalinux8, packageManager: dnf, target: gfx1100 }
-      #- { os: almalinux8, packageManager: dnf, target: gfx1030 }
+      - { os: almalinux8, packageManager: dnf, target: gfx1030 }
    testJobs:
      - { os: ubuntu2204, packageManager: apt, target: gfx942 }
      - { os: ubuntu2204, packageManager: apt, target: gfx90a }
--- a/.azuredevops/components/rocSPARSE.yml
+++ b/.azuredevops/components/rocSPARSE.yml
@@ -73,7 +73,7 @@ parameters:
      - { os: ubuntu2204, packageManager: apt, target: gfx90a }
      - { os: ubuntu2204, packageManager: apt, target: gfx1201 }
      - { os: ubuntu2204, packageManager: apt, target: gfx1100 }
-      #- { os: ubuntu2204, packageManager: apt, target: gfx1030 }
+      - { os: ubuntu2204, packageManager: apt, target: gfx1030 }
    testJobs:
      - { os: ubuntu2204, packageManager: apt, target: gfx942 }
      - { os: ubuntu2204, packageManager: apt, target: gfx90a }
--- a/.azuredevops/components/rocWMMA.yml
+++ b/.azuredevops/components/rocWMMA.yml
@@ -70,7 +70,7 @@ jobs:
    variables:
    - group: common
    - template: /.azuredevops/variables-global.yml
-    pool: ${{ variables.MEDIUM_BUILD_POOL }}
+    pool: ${{ variables.HIGH_BUILD_POOL }}
    workspace:
      clean: all
    steps:
--- a/.azuredevops/components/rocm-libraries.yml
+++ b/.azuredevops/components/rocm-libraries.yml
@@ -36,10 +36,8 @@ parameters:
    - gfortran
    - git
    - libdrm-dev
-    - liblapack-dev
    - libmsgpack-dev
    - libnuma-dev
-    - libopenblas-dev
    - ninja-build
    - python3-pip
    - python3-venv
@@ -48,8 +46,6 @@ parameters:
  default:
    - joblib
    - "packaging>=22.0"
-    - pytest
-    - pytest-cmake
    - --upgrade
 - name: rocmDependencies
  type: object
@@ -102,12 +98,12 @@ jobs:
    workspace:
      clean: all
    steps:
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-cmake-latest.yml
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-other.yml
      parameters:
        aptPackages: ${{ parameters.aptPackages }}
        pipModules: ${{ parameters.pipModules }}
        packageManager: ${{ job.packageManager }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-cmake-latest.yml
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
      parameters:
@@ -138,26 +134,12 @@ jobs:
          rocm-libraries | ${{ job.os }} | ${{ job.target }} | $(DAY_STRING)
          rocm-libraries | ${{ job.os }} | ${{ job.target }}
          rocm-libraries | ${{ job.os }}
-    - task: Bash@3
-      displayName: Add paths for CMake and Python site-packages binaries
-      inputs:
-        targetType: inline
-        script: |
-          USER_BASE=$(python3 -m site --user-base)
-          echo "##vso[task.prependpath]$USER_BASE/bin"
-          echo "##vso[task.setvariable variable=PytestCmakePath]$USER_BASE/share/Pytest/cmake"
-        displayName: Set cmake configure paths
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/build-cmake.yml
      parameters:
        os: ${{ job.os }}
        extraBuildFlags: >-
-          -D CMAKE_PREFIX_PATH=$(Agent.BuildDirectory)/rocm;$(Agent.BuildDirectory)/vendor;$(PytestCmakePath)
-          -D CMAKE_INCLUDE_PATH=$(Agent.BuildDirectory)/rocm/llvm/include
-          -D CMAKE_CXX_COMPILER=$(Agent.BuildDirectory)/rocm/llvm/bin/amdclang++
-          -D CMAKE_C_COMPILER=$(Agent.BuildDirectory)/rocm/llvm/bin/amdclang
-          -D CMAKE_CXX_COMPILER_LAUNCHER=ccache
-          -D CMAKE_C_COMPILER_LAUNCHER=ccache
-          -G Ninja
+          -DROCM_LIBRARIES_SUPERBUILD=ON
+          -GNinja
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/manifest.yml
      parameters:
        componentName: ${{ parameters.componentName }}
--- a/.azuredevops/components/rocprofiler-compute.yml
+++ b/.azuredevops/components/rocprofiler-compute.yml
@@ -1,29 +1,10 @@
 parameters:
- name: componentName
-  type: string
-  default: rocprofiler-compute
 - name: checkoutRepo
  type: string
  default: 'self'
 - name: checkoutRef
  type: string
  default: ''
-# monorepo related parameters
- name: sparseCheckoutDir
-  type: string
-  default: ''
- name: triggerDownstreamJobs
-  type: boolean
-  default: false
- name: downstreamAggregateNames
-  type: string
-  default: ''
- name: buildDependsOn
-  type: object
-  default: null
- name: unifiedBuild
-  type: boolean
-  default: false
 # set to true if doing full build of ROCm stack
 # and dependencies are pulled from same pipeline
 - name: aggregatePipeline
@@ -97,10 +78,6 @@ parameters:
 jobs:
 - ${{ each job in parameters.jobMatrix.buildJobs }}:
  - job: rocprofiler_compute_build_${{ job.target }}
-    ${{ if parameters.buildDependsOn }}:
-      dependsOn:
-        - ${{ each build in parameters.buildDependsOn }}:
-          - ${{ build }}_${{ job.os }}_${{ job.target }}
    variables:
    - group: common
    - template: /.azuredevops/variables-global.yml
@@ -117,19 +94,15 @@ jobs:
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
      parameters:
        checkoutRepo: ${{ parameters.checkoutRepo }}
-        sparseCheckoutDir: ${{ parameters.sparseCheckoutDir }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/build-cmake.yml
      parameters:
        extraBuildFlags: >-
          -GNinja
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/manifest.yml
      parameters:
-        componentName: ${{ parameters.componentName }}
-        sparseCheckoutDir: ${{ parameters.sparseCheckoutDir }}
        gpuTarget: ${{ job.target }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-upload.yml
      parameters:
-        componentName: ${{ parameters.componentName }}
        gpuTarget: ${{ job.target }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-links.yml
    # - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/docker-container.yml
@@ -138,83 +111,78 @@ jobs:
    #     pipModules: ${{ parameters.pipModules }}
    #     gpuTarget: ${{ job.target }}

- ${{ if eq(parameters.unifiedBuild, False) }}:
-  - ${{ each job in parameters.jobMatrix.testJobs }}:
-    - job: rocprofiler_compute_test_${{ job.target }}
-      timeoutInMinutes: 120
-      dependsOn: rocprofiler_compute_build_${{ job.target }}
-      condition:
-        and(succeeded(),
-          eq(variables['ENABLE_${{ upper(job.target) }}_TESTS'], 'true'),
-          not(containsValue(split(variables['DISABLED_${{ upper(job.target) }}_TESTS'], ','), '${{ parameters.componentName }}')),
-          eq(${{ parameters.aggregatePipeline }}, False)
-        )
-      variables:
-      - group: common
-      - template: /.azuredevops/variables-global.yml
-      - name: PYTHON_VERSION
-        value: 3.10
-      pool: ${{ job.target }}_test_pool
-      workspace:
-        clean: all
-      steps:
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-other.yml
-        parameters:
-          aptPackages: ${{ parameters.aptPackages }}
-          pipModules: ${{ parameters.pipModules }}
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
-        parameters:
-          checkoutRepo: ${{ parameters.checkoutRepo }}
-          sparseCheckoutDir: ${{ parameters.sparseCheckoutDir }}
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/local-artifact-download.yml
-        parameters:
-          preTargetFilter: ${{ parameters.componentName }}
-          gpuTarget: ${{ job.target }}
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-aqlprofile.yml
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
-        parameters:
-          checkoutRef: ${{ parameters.checkoutRef }}
-          dependencyList: ${{ parameters.rocmTestDependencies }}
-          gpuTarget: ${{ job.target }}
-          ${{ if parameters.triggerDownstreamJobs }}:
-            downstreamAggregateNames: ${{ parameters.downstreamAggregateNames }}
-      - task: Bash@3
-        displayName: Add en_US.UTF-8 locale
-        inputs:
-          targetType: inline
-          script: |
-            sudo locale-gen en_US.UTF-8
-            sudo update-locale
-            locale -a
-      - task: Bash@3
-        displayName: Add ROCm binaries to PATH
-        inputs:
-          targetType: inline
-          script: |
-            echo "##vso[task.prependpath]$(Agent.BuildDirectory)/rocm/bin"
-            echo "##vso[task.prependpath]$(Agent.BuildDirectory)/rocm/llvm/bin"
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/build-cmake.yml
-        parameters:
-          extraBuildFlags: >-
-            -DCMAKE_HIP_ARCHITECTURES=${{ job.target }}
-            -DCMAKE_C_COMPILER=$(Agent.BuildDirectory)/rocm/llvm/bin/amdclang
-            -DCMAKE_MODULE_PATH=$(Agent.BuildDirectory)/rocm/lib/cmake/hip
-            -DCMAKE_PREFIX_PATH=$(Agent.BuildDirectory)/rocm
-            -DROCM_PATH=$(Agent.BuildDirectory)/rocm
-            -DCMAKE_BUILD_TYPE=Release
-            -DENABLE_TESTS=ON
-            -DINSTALL_TESTS=ON
-            -GNinja
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/gpu-diagnostics.yml
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/test.yml
-        parameters:
-          componentName: ${{ parameters.componentName }}
-          testDir: $(Build.BinariesDirectory)/libexec/rocprofiler-compute
-          testExecutable: ROCM_PATH=$(Agent.BuildDirectory)/rocm ctest
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/docker-container.yml
-        parameters:
-          aptPackages: ${{ parameters.aptPackages }}
-          pipModules: ${{ parameters.pipModules }}
-          environment: test
-          gpuTarget: ${{ job.target }}
+- ${{ each job in parameters.jobMatrix.testJobs }}:
+  - job: rocprofiler_compute_test_${{ job.target }}
+    timeoutInMinutes: 120
+    dependsOn: rocprofiler_compute_build_${{ job.target }}
+    condition:
+      and(succeeded(),
+        eq(variables['ENABLE_${{ upper(job.target) }}_TESTS'], 'true'),
+        not(containsValue(split(variables['DISABLED_${{ upper(job.target) }}_TESTS'], ','), variables['Build.DefinitionName'])),
+        eq(${{ parameters.aggregatePipeline }}, False)
+      )
+    variables:
+    - group: common
+    - template: /.azuredevops/variables-global.yml
+    - name: PYTHON_VERSION
+      value: 3.10
+    pool: ${{ job.target }}_test_pool
+    workspace:
+      clean: all
+    steps:
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-other.yml
+      parameters:
+        aptPackages: ${{ parameters.aptPackages }}
+        pipModules: ${{ parameters.pipModules }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
+      parameters:
+        checkoutRepo: ${{ parameters.checkoutRepo }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/local-artifact-download.yml
+      parameters:
+        gpuTarget: ${{ job.target }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-aqlprofile.yml
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
+      parameters:
+        checkoutRef: ${{ parameters.checkoutRef }}
+        dependencyList: ${{ parameters.rocmTestDependencies }}
+        gpuTarget: ${{ job.target }}
+    - task: Bash@3
+      displayName: Add en_US.UTF-8 locale
+      inputs:
+        targetType: inline
+        script: |
+          sudo locale-gen en_US.UTF-8
+          sudo update-locale
+          locale -a
+    - task: Bash@3
+      displayName: Add ROCm binaries to PATH
+      inputs:
+        targetType: inline
+        script: |
+          echo "##vso[task.prependpath]$(Agent.BuildDirectory)/rocm/bin"
+          echo "##vso[task.prependpath]$(Agent.BuildDirectory)/rocm/llvm/bin"
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/build-cmake.yml
+      parameters:
+        extraBuildFlags: >-
+          -DCMAKE_HIP_ARCHITECTURES=${{ job.target }}
+          -DCMAKE_C_COMPILER=$(Agent.BuildDirectory)/rocm/llvm/bin/amdclang
+          -DCMAKE_MODULE_PATH=$(Agent.BuildDirectory)/rocm/lib/cmake/hip
+          -DCMAKE_PREFIX_PATH=$(Agent.BuildDirectory)/rocm
+          -DROCM_PATH=$(Agent.BuildDirectory)/rocm
+          -DCMAKE_BUILD_TYPE=Release
+          -DENABLE_TESTS=ON
+          -DINSTALL_TESTS=ON
+          -GNinja
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/gpu-diagnostics.yml
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/test.yml
+      parameters:
+        componentName: rocprofiler-compute
+        testDir: $(Build.BinariesDirectory)/libexec/rocprofiler-compute
+        testExecutable: ROCM_PATH=$(Agent.BuildDirectory)/rocm ctest
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/docker-container.yml
+      parameters:
+        aptPackages: ${{ parameters.aptPackages }}
+        pipModules: ${{ parameters.pipModules }}
+        environment: test
+        gpuTarget: ${{ job.target }}
--- a/.azuredevops/components/rocprofiler.yml
+++ b/.azuredevops/components/rocprofiler.yml
@@ -8,22 +8,6 @@ parameters:
 - name: checkoutRef
  type: string
  default: ''
-# monorepo related parameters
- name: sparseCheckoutDir
-  type: string
-  default: ''
- name: triggerDownstreamJobs
-  type: boolean
-  default: false
- name: downstreamAggregateNames
-  type: string
-  default: ''
- name: buildDependsOn
-  type: object
-  default: null
- name: unifiedBuild
-  type: boolean
-  default: false
 # set to true if doing full build of ROCm stack
 # and dependencies are pulled from same pipeline
 - name: aggregatePipeline
@@ -86,10 +70,6 @@ parameters:
 jobs:
 - ${{ each job in parameters.jobMatrix.buildJobs }}:
  - job: ${{ parameters.componentName }}_build_${{ job.os }}_${{ job.target }}
-    ${{ if parameters.buildDependsOn }}:
-      dependsOn:
-        - ${{ each build in parameters.buildDependsOn }}:
-          - ${{ build }}_${{ job.os }}_${{ job.target }}
    variables:
    - group: common
    - template: /.azuredevops/variables-global.yml
@@ -114,7 +94,6 @@ jobs:
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
      parameters:
        checkoutRepo: ${{ parameters.checkoutRepo }}
-        sparseCheckoutDir: ${{ parameters.sparseCheckoutDir }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-vendor.yml
      parameters:
        dependencyList:
@@ -129,8 +108,6 @@ jobs:
        gpuTarget: ${{ job.target }}
        os: ${{ job.os }}
        aggregatePipeline: ${{ parameters.aggregatePipeline }}
-        ${{ if parameters.triggerDownstreamJobs }}:
-          downstreamAggregateNames: ${{ parameters.downstreamAggregateNames }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/build-cmake.yml
      parameters:
        os: ${{ job.os }}
@@ -138,7 +115,6 @@ jobs:
        extraBuildFlags: >-
          -DCMAKE_MODULE_PATH=$(Build.SourcesDirectory)/cmake_modules;$(Agent.BuildDirectory)/rocm/lib/cmake;$(Agent.BuildDirectory)/rocm/lib/cmake/hip;$(Agent.BuildDirectory)/rocm/lib64/cmake;$(Agent.BuildDirectory)/rocm/lib64/cmake/hip
          -DCMAKE_PREFIX_PATH="$(Agent.BuildDirectory)/rocm;$(Agent.BuildDirectory)/vendor"
-          -DROCM_PATH=$(Agent.BuildDirectory)/rocm
          -DCMAKE_POSITION_INDEPENDENT_CODE=ON
          -DENABLE_LDCONFIG=OFF
          -DUSE_PROF_API=1
@@ -146,13 +122,10 @@ jobs:
        multithreadFlag: -- -j32
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/manifest.yml
      parameters:
-        componentName: ${{ parameters.componentName }}
        gpuTarget: ${{ job.target }}
        os: ${{ job.os }}
-        sparseCheckoutDir: ${{ parameters.sparseCheckoutDir }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-upload.yml
      parameters:
-        componentName: ${{ parameters.componentName }}
        gpuTarget: ${{ job.target }}
        os: ${{ job.os }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-links.yml
@@ -166,68 +139,63 @@ jobs:
            - HIP_ROCCLR_HOME:::/home/user/workspace/rocm
            - ROCM_PATH:::/home/user/workspace/rocm

- ${{ if eq(parameters.unifiedBuild, False) }}:
-  - ${{ each job in parameters.jobMatrix.testJobs }}:
-    - job: ${{ parameters.componentName }}_test_${{ job.os }}_${{ job.target }}
-      dependsOn: ${{ parameters.componentName }}_build_${{ job.os }}_${{ job.target }}
-      condition:
-        and(succeeded(),
-          eq(variables['ENABLE_${{ upper(job.target) }}_TESTS'], 'true'),
-          not(containsValue(split(variables['DISABLED_${{ upper(job.target) }}_TESTS'], ','), variables['Build.DefinitionName'])),
-          eq(${{ parameters.aggregatePipeline }}, False)
-        )
-      variables:
-      - group: common
-      - template: /.azuredevops/variables-global.yml
-      - name: ROCM_PATH
-        value: $(Agent.BuildDirectory)/rocm
-      - name: LD_LIBRARY_PATH
-        value: $(Agent.BuildDirectory)/rocm/lib/rocprofiler:$(Agent.BuildDirectory)/rocm/share/rocprofiler/tests-v1/test:$(Agent.BuildDirectory)/rocm/share/rocprofiler/tests
-      pool: ${{ job.target }}_test_pool
-      workspace:
-        clean: all
-      steps:
-      - checkout: none
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-other.yml
-        parameters:
-          aptPackages: ${{ parameters.aptPackages }}
-          packageManager: ${{ job.packageManager }}
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/local-artifact-download.yml
-        parameters:
-          preTargetFilter: ${{ parameters.componentName }}
-          gpuTarget: ${{ job.target }}
-          os: ${{ job.os }}
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-aqlprofile.yml
-        parameters:
-          os: ${{ job.os }}
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
-        parameters:
-          checkoutRef: ${{ parameters.checkoutRef }}
-          dependencyList: ${{ parameters.rocmDependencies }}
-          gpuTarget: ${{ job.target }}
-          os: ${{ job.os }}
-          ${{ if parameters.triggerDownstreamJobs }}:
-            downstreamAggregateNames: ${{ parameters.downstreamAggregateNames }}
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/gpu-diagnostics.yml
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/test.yml
-        parameters:
-          componentName: rocprofilerV1
-          testDir: $(Agent.BuildDirectory)/rocm/share/rocprofiler/tests-v1
-          testExecutable:  ./run.sh
-          testParameters: ''
-          testPublishResults: false
-          os: ${{ job.os }}
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/test.yml
-        parameters:
-          componentName: rocprofilerV2
-          testDir: $(Agent.BuildDirectory)/rocm
-          testExecutable:  share/rocprofiler/tests/runUnitTests
-          testParameters: '--gtest_output=xml:./test_output.xml --gtest_color=yes'
-          os: ${{ job.os }}
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/docker-container.yml
-        parameters:
-          aptPackages: ${{ parameters.aptPackages }}
-          pipModules: ${{ parameters.pipModules }}
-          environment: test
-          gpuTarget: ${{ job.target }}
+- ${{ each job in parameters.jobMatrix.testJobs }}:
+  - job: ${{ parameters.componentName }}_test_${{ job.os }}_${{ job.target }}
+    dependsOn: ${{ parameters.componentName }}_build_${{ job.os }}_${{ job.target }}
+    condition:
+      and(succeeded(),
+        eq(variables['ENABLE_${{ upper(job.target) }}_TESTS'], 'true'),
+        not(containsValue(split(variables['DISABLED_${{ upper(job.target) }}_TESTS'], ','), variables['Build.DefinitionName'])),
+        eq(${{ parameters.aggregatePipeline }}, False)
+      )
+    variables:
+    - group: common
+    - template: /.azuredevops/variables-global.yml
+    - name: ROCM_PATH
+      value: $(Agent.BuildDirectory)/rocm
+    - name: LD_LIBRARY_PATH
+      value: $(Agent.BuildDirectory)/rocm/lib/rocprofiler:$(Agent.BuildDirectory)/rocm/share/rocprofiler/tests-v1/test:$(Agent.BuildDirectory)/rocm/share/rocprofiler/tests
+    pool: ${{ job.target }}_test_pool
+    workspace:
+      clean: all
+    steps:
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-other.yml
+      parameters:
+        aptPackages: ${{ parameters.aptPackages }}
+        packageManager: ${{ job.packageManager }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/local-artifact-download.yml
+      parameters:
+        gpuTarget: ${{ job.target }}
+        os: ${{ job.os }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-aqlprofile.yml
+      parameters:
+        os: ${{ job.os }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
+      parameters:
+        checkoutRef: ${{ parameters.checkoutRef }}
+        dependencyList: ${{ parameters.rocmDependencies }}
+        gpuTarget: ${{ job.target }}
+        os: ${{ job.os }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/gpu-diagnostics.yml
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/test.yml
+      parameters:
+        componentName: rocprofilerV1
+        testDir: $(Agent.BuildDirectory)/rocm/share/rocprofiler/tests-v1
+        testExecutable:  ./run.sh
+        testParameters: ''
+        testPublishResults: false
+        os: ${{ job.os }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/test.yml
+      parameters:
+        componentName: rocprofilerV2
+        testDir: $(Agent.BuildDirectory)/rocm
+        testExecutable:  share/rocprofiler/tests/runUnitTests
+        testParameters: '--gtest_output=xml:./test_output.xml --gtest_color=yes'
+        os: ${{ job.os }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/docker-container.yml
+      parameters:
+        aptPackages: ${{ parameters.aptPackages }}
+        pipModules: ${{ parameters.pipModules }}
+        environment: test
+        gpuTarget: ${{ job.target }}
--- a/.azuredevops/components/roctracer.yml
+++ b/.azuredevops/components/roctracer.yml
@@ -8,22 +8,6 @@ parameters:
 - name: checkoutRef
  type: string
  default: ''
-# monorepo related parameters
- name: sparseCheckoutDir
-  type: string
-  default: ''
- name: triggerDownstreamJobs
-  type: boolean
-  default: false
- name: downstreamAggregateNames
-  type: string
-  default: ''
- name: buildDependsOn
-  type: object
-  default: null
- name: unifiedBuild
-  type: boolean
-  default: false
 # set to true if doing full build of ROCm stack
 # and dependencies are pulled from same pipeline
 - name: aggregatePipeline
@@ -81,10 +65,6 @@ parameters:
 jobs:
 - ${{ each job in parameters.jobMatrix.buildJobs }}:
  - job: ${{ parameters.componentName }}_build_${{ job.os }}_${{ job.target }}
-    ${{ if parameters.buildDependsOn }}:
-      dependsOn:
-        - ${{ each build in parameters.buildDependsOn }}:
-          - ${{ build }}_${{ job.os }}_${{ job.target }}
    variables:
    - group: common
    - template: /.azuredevops/variables-global.yml
@@ -107,7 +87,6 @@ jobs:
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
      parameters:
        checkoutRepo: ${{ parameters.checkoutRepo }}
-        sparseCheckoutDir: ${{ parameters.sparseCheckoutDir }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
      parameters:
        checkoutRef: ${{ parameters.checkoutRef }}
@@ -115,8 +94,6 @@ jobs:
        gpuTarget: ${{ job.target }}
        aggregatePipeline: ${{ parameters.aggregatePipeline }}
        os: ${{ job.os }}
-        ${{ if parameters.triggerDownstreamJobs }}:
-          downstreamAggregateNames: ${{ parameters.downstreamAggregateNames }}
    # the linker flags will not affect ubuntu2204 builds as the paths do not exist
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/build-cmake.yml
      parameters:
@@ -132,13 +109,10 @@ jobs:
          -GNinja
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/manifest.yml
      parameters:
-        componentName: ${{ parameters.componentName }}
-        sparseCheckoutDir: ${{ parameters.sparseCheckoutDir }}
        os: ${{ job.os }}
        gpuTarget: ${{ job.target }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-upload.yml
      parameters:
-        componentName: ${{ parameters.componentName }}
        os: ${{ job.os }}
        gpuTarget: ${{ job.target }}
    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-links.yml
@@ -149,57 +123,53 @@ jobs:
    #     gpuTarget: ${{ job.target }}
    #     registerROCmPackages: true

- ${{ if eq(parameters.unifiedBuild, False) }}:
-  - ${{ each job in parameters.jobMatrix.testJobs }}:
-    - job: ${{ parameters.componentName }}_test_${{ job.os }}_${{ job.target }}
-      dependsOn: ${{ parameters.componentName }}_build_${{ job.os }}_${{ job.target }}
-      condition:
-        and(succeeded(),
-          eq(variables['ENABLE_${{ upper(job.target) }}_TESTS'], 'true'),
-          not(containsValue(split(variables['DISABLED_${{ upper(job.target) }}_TESTS'], ','), '${{ parameters.componentName }}')),
-          eq(${{ parameters.aggregatePipeline }}, False)
-        )
-      variables:
-      - group: common
-      - template: /.azuredevops/variables-global.yml
-      pool: ${{ job.target }}_test_pool
-      workspace:
-        clean: all
-      steps:
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-other.yml
-        parameters:
-          aptPackages: ${{ parameters.aptPackages }}
-          pipModules: ${{ parameters.pipModules }}
-          packageManager: ${{ job.packageManager }}
-          registerROCmPackages: true
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/local-artifact-download.yml
-        parameters:
-          preTargetFilter: ${{ parameters.componentName }}
-          gpuTarget: ${{ job.target }}
-          os: ${{ job.os }}
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-aqlprofile.yml
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
-        parameters:
-          checkoutRef: ${{ parameters.checkoutRef }}
-          dependencyList: ${{ parameters.rocmTestDependencies }}
-          gpuTarget: ${{ job.target }}
-          os: ${{ job.os }}
-          ${{ if parameters.triggerDownstreamJobs }}:
-            downstreamAggregateNames: ${{ parameters.downstreamAggregateNames }}
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/gpu-diagnostics.yml
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/test.yml
-        parameters:
-          componentName: ${{ parameters.componentName }}
-          testExecutable: $(Agent.BuildDirectory)/rocm/share/roctracer/run_tests.sh
-          testParameters: ''
-          testDir: $(Agent.BuildDirectory)
-          testPublishResults: false
-          os: ${{ job.os }}
-      - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/docker-container.yml
-        parameters:
-          aptPackages: ${{ parameters.aptPackages }}
-          pipModules: ${{ parameters.pipModules }}
-          environment: test
-          gpuTarget: ${{ job.target }}
-          registerROCmPackages: true
+- ${{ each job in parameters.jobMatrix.testJobs }}:
+  - job: ${{ parameters.componentName }}_test_${{ job.os }}_${{ job.target }}
+    dependsOn: ${{ parameters.componentName }}_build_${{ job.os }}_${{ job.target }}
+    condition:
+      and(succeeded(),
+        eq(variables['ENABLE_${{ upper(job.target) }}_TESTS'], 'true'),
+        not(containsValue(split(variables['DISABLED_${{ upper(job.target) }}_TESTS'], ','), variables['Build.DefinitionName'])),
+        eq(${{ parameters.aggregatePipeline }}, False)
+      )
+    variables:
+    - group: common
+    - template: /.azuredevops/variables-global.yml
+    pool: ${{ job.target }}_test_pool
+    workspace:
+      clean: all
+    steps:
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-other.yml
+      parameters:
+        aptPackages: ${{ parameters.aptPackages }}
+        pipModules: ${{ parameters.pipModules }}
+        packageManager: ${{ job.packageManager }}
+        registerROCmPackages: true
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/local-artifact-download.yml
+      parameters:
+        gpuTarget: ${{ job.target }}
+        os: ${{ job.os }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-aqlprofile.yml
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
+      parameters:
+        checkoutRef: ${{ parameters.checkoutRef }}
+        dependencyList: ${{ parameters.rocmTestDependencies }}
+        gpuTarget: ${{ job.target }}
+        os: ${{ job.os }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/gpu-diagnostics.yml
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/test.yml
+      parameters:
+        componentName: roctracer
+        testExecutable: $(Agent.BuildDirectory)/rocm/share/roctracer/run_tests.sh
+        testParameters: ''
+        testDir: $(Agent.BuildDirectory)
+        testPublishResults: false
+        os: ${{ job.os }}
+    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/docker-container.yml
+      parameters:
+        aptPackages: ${{ parameters.aptPackages }}
+        pipModules: ${{ parameters.pipModules }}
+        environment: test
+        gpuTarget: ${{ job.target }}
+        registerROCmPackages: true
--- a/.azuredevops/dependencies/fmtlib.yml
+++ b/.azuredevops/dependencies/fmtlib.yml
@@ -1,67 +0,0 @@
-parameters:
- name: checkoutRepo
-  type: string
-  default: 'self'
- name: checkoutRef
-  type: string
-  default: ''
- name: fmtlibVersion
-  type: string
-  default: ''
- name: aptPackages
-  type: object
-  default:
-    - cmake
-    - git
-    - ninja-build
-    - libfmt-dev
-
- name: jobMatrix
-  type: object
-  default:
-    buildJobs:
-      - { os: ubuntu2204, packageManager: apt}
-      - { os: almalinux8, packageManager: dnf}
-
-jobs:
- ${{ each job in parameters.jobMatrix.buildJobs }}:
-  - job: fmtlib_${{ job.os }}
-    variables:
-    - group: common
-    - template: /.azuredevops/variables-global.yml
-    pool:
-      vmImage: 'ubuntu-22.04'
-    ${{ if eq(job.os, 'almalinux8') }}:
-      container:
-        image: rocmexternalcicd.azurecr.io/manylinux228:latest
-        endpoint: ContainerService3
-    workspace:
-      clean: all
-    steps:
-    - checkout: none
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-other.yml
-      parameters:
-        aptPackages: ${{ parameters.aptPackages }}
-        packageManager: ${{ job.packageManager }}
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
-    - task: Bash@3
-      displayName: Clone fmtlib ${{ parameters.fmtlibVersion }}
-      inputs:
-        targetType: inline
-        script: git clone https://github.com/fmtlib/fmt.git -b ${{ parameters.fmtlibVersion }}
-        workingDirectory: $(Agent.BuildDirectory)
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/build-cmake.yml
-      parameters:
-        os: ${{ job.os }}
-        cmakeBuildDir: $(Agent.BuildDirectory)/fmt/build
-        cmakeSourceDir: $(Agent.BuildDirectory)/fmt
-        useAmdclang: false
-        extraBuildFlags: >-
-          -DCMAKE_BUILD_TYPE=Release
-          -DFMT_SYSTEM_HEADERS=ON
-          -DFMT_INSTALL=ON
-          -DFMT_TEST=OFF
-          -GNinja
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-upload.yml
-      parameters:
-        os: ${{ job.os }}
--- a/.azuredevops/dependencies/spdlog.yml
+++ b/.azuredevops/dependencies/spdlog.yml
@@ -1,71 +0,0 @@
-parameters:
- name: checkoutRepo
-  type: string
-  default: 'self'
- name: checkoutRef
-  type: string
-  default: ''
- name: spdlogVersion
-  type: string
-  default: ''
- name: aptPackages
-  type: object
-  default:
-    - cmake
-    - git
-    - ninja-build
-
- name: jobMatrix
-  type: object
-  default:
-    buildJobs:
-      - { os: ubuntu2204, packageManager: apt}
-      - { os: almalinux8, packageManager: dnf}
-
-jobs:
- ${{ each job in parameters.jobMatrix.buildJobs }}:
-  - job: spdlog_${{ job.os }}
-    variables:
-    - group: common
-    - template: /.azuredevops/variables-global.yml
-    pool:
-      vmImage: 'ubuntu-22.04'
-    ${{ if eq(job.os, 'almalinux8') }}:
-      container:
-        image: rocmexternalcicd.azurecr.io/manylinux228:latest
-        endpoint: ContainerService3
-    workspace:
-      clean: all
-    steps:
-    - checkout: none
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-other.yml
-      parameters:
-        aptPackages: ${{ parameters.aptPackages }}
-        packageManager: ${{ job.packageManager }}
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-vendor.yml
-      parameters:
-        dependencyList:
-          - fmtlib
-    - task: Bash@3
-      displayName: Clone spdlog ${{ parameters.spdlogVersion }}
-      inputs:
-        targetType: inline
-        script: git clone https://github.com/gabime/spdlog.git -b ${{ parameters.spdlogVersion }}
-        workingDirectory: $(Agent.BuildDirectory)
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/build-cmake.yml
-      parameters:
-        os: ${{ job.os }}
-        cmakeBuildDir: $(Agent.BuildDirectory)/spdlog/build
-        cmakeSourceDir: $(Agent.BuildDirectory)/spdlog
-        useAmdclang: false
-        extraBuildFlags: >-
-          -DCMAKE_PREFIX_PATH=$(Agent.BuildDirectory)/vendor
-          -DCMAKE_BUILD_TYPE=Release
-          -DSPDLOG_USE_STD_FORMAT=OFF
-          -DSPDLOG_FMT_EXTERNAL_HO=ON
-          -DSPDLOG_INSTALL=ON
-          -GNinja
-    - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-upload.yml
-      parameters:
-        os: ${{ job.os }}
--- a/.azuredevops/nightly/pytorch.yml
+++ b/.azuredevops/nightly/pytorch.yml
@@ -397,7 +397,6 @@ jobs:
  - template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
  - task: DownloadPipelineArtifact@2
    displayName: 'Download Pipeline Wheel Files'
-    retryCountOnTaskFailure: 3
    inputs:
      itemPattern: '**/*.whl'
      targetPath: $(Agent.BuildDirectory)
--- a/.azuredevops/nightly/rocm-nightly.yml
+++ b/.azuredevops/nightly/rocm-nightly.yml
@@ -93,7 +93,7 @@ schedules:
 jobs:
 - ${{ each job in parameters.jobList }}:
  - job: nightly_${{ job.os }}_${{ job.target }}
-    timeoutInMinutes: 120
+    timeoutInMinutes: 90
    variables:
    - group: common
    - template: /.azuredevops/variables-global.yml
@@ -226,7 +226,6 @@ jobs:
            cat Dockerfile
    - task: Docker@2
      displayName: Build and upload Docker image
-      retryCountOnTaskFailure: 3
      inputs:
        containerRegistry: ContainerService3
        repository: 'nightly-${{ job.os }}-${{ job.target }}'
--- a/.azuredevops/tag-builds/fmtlib.yml
+++ b/.azuredevops/tag-builds/fmtlib.yml
@@ -1,23 +0,0 @@
-variables:
- group: common
- template: /.azuredevops/variables-global.yml
-
-parameters:
- name: fmtlibVersion
-  type: string
-  default: "11.1.3"
-
-resources:
-  repositories:
-  - repository: pipelines_repo
-    type: github
-    endpoint: ROCm
-    name: ROCm/ROCm
-
-trigger: none
-pr: none
-
-jobs:
-  - template: ${{ variables.CI_DEPENDENCIES_PATH }}/fmtlib.yml
-    parameters:
-      fmtlibVersion: ${{ parameters.fmtlibVersion }}
--- a/.azuredevops/tag-builds/spdlog.yml
+++ b/.azuredevops/tag-builds/spdlog.yml
@@ -1,23 +0,0 @@
-variables:
- group: common
- template: /.azuredevops/variables-global.yml
-
-parameters:
- name: spdlogVersion
-  type: string
-  default: "v1.15.1"
-
-resources:
-  repositories:
-  - repository: pipelines_repo
-    type: github
-    endpoint: ROCm
-    name: ROCm/ROCm
-
-trigger: none
-pr: none
-
-jobs:
-  - template: ${{ variables.CI_DEPENDENCIES_PATH }}/spdlog.yml
-    parameters:
-      spdlogVersion: ${{ parameters.spdlogVersion }}
--- a/.azuredevops/templates/steps/artifact-download.yml
+++ b/.azuredevops/templates/steps/artifact-download.yml
@@ -24,12 +24,8 @@ parameters:
 steps:
 - task: DownloadPipelineArtifact@2
  displayName: Download ${{ parameters.componentName }}
-  retryCountOnTaskFailure: 3
  inputs:
-    ${{ if eq(parameters.componentName, 'clr') }}:
-      itemPattern: '**/*${{ parameters.componentName }}*${{ parameters.fileFilter }}*amd*' # filter out nvidia clr artifacts
-    ${{ else }}:
-      itemPattern: '**/*${{ parameters.componentName }}*${{ parameters.fileFilter }}*'
+    itemPattern: '**/*${{ parameters.componentName }}*${{ parameters.fileFilter }}*'
    targetPath: '$(Pipeline.Workspace)/d'
    allowPartiallySucceededBuilds: true
    ${{ if parameters.aggregatePipeline }}:
--- a/.azuredevops/templates/steps/dependencies-apt.yml
+++ b/.azuredevops/templates/steps/dependencies-apt.yml
@@ -10,7 +10,6 @@ steps:
 - ${{ if eq(parameters.registerROCmPackages, true) }}:
  - task: Bash@3
    displayName: 'Register AMDGPU & ROCm repos (apt)'
-    retryCountOnTaskFailure: 3
    inputs:
      targetType: inline
      script: |
@@ -21,8 +20,7 @@ steps:
        echo -e 'Package: *\nPin: release o=repo.radeon.com\nPin-Priority: 600' | sudo tee /etc/apt/preferences.d/rocm-pin-600
        sudo apt update
 - task: Bash@3
-  displayName: 'APT update and install packages'
-  retryCountOnTaskFailure: 3
+  displayName: 'sudo apt-get update'
  inputs:
    targetType: inline
    script: |
@@ -30,6 +28,15 @@ steps:
      echo "deb http://archive.ubuntu.com/ubuntu/ jammy-updates main restricted universe multiverse" | sudo tee -a /etc/apt/sources.list.d/default.list
      echo "deb http://archive.ubuntu.com/ubuntu/ jammy-backports main restricted universe multiverse" | sudo tee -a /etc/apt/sources.list.d/default.list
      echo "deb http://archive.ubuntu.com/ubuntu/ jammy-security main restricted universe multiverse" | sudo tee -a /etc/apt/sources.list.d/default.list
-      sudo DEBIAN_FRONTEND=noninteractive apt-get --yes update && \
-        sudo DEBIAN_FRONTEND=noninteractive apt-get --yes --fix-broken install && \
-        sudo DEBIAN_FRONTEND=noninteractive apt-get --yes --fix-missing install ${{ join(' ', parameters.aptPackages) }}
+      sudo DEBIAN_FRONTEND=noninteractive apt-get --yes update
+- task: Bash@3
+  displayName: 'sudo apt-get fix'
+  inputs:
+    targetType: inline
+    script: sudo DEBIAN_FRONTEND=noninteractive apt-get --yes --fix-broken install
+- ${{ if gt(length(parameters.aptPackages), 0) }}:
+  - task: Bash@3
+    displayName: 'sudo apt-get install ...'
+    inputs:
+      targetType: inline
+      script: sudo DEBIAN_FRONTEND=noninteractive apt-get --yes --fix-missing install ${{ join(' ', parameters.aptPackages) }}
--- a/.azuredevops/templates/steps/dependencies-aqlprofile.yml
+++ b/.azuredevops/templates/steps/dependencies-aqlprofile.yml
@@ -5,28 +5,51 @@ parameters:

 steps:
 - task: Bash@3
-  displayName: Download and install aqlprofile
-  retryCountOnTaskFailure: 3
+  displayName: Get aqlprofile package name
+  inputs:
+    targetType: inline
+    ${{ if eq(parameters.os, 'ubuntu2204') }}:
+      script: |
+        export packageName=$(curl -s https://repo.radeon.com/rocm/apt/$(REPO_RADEON_VERSION)/pool/main/h/hsa-amd-aqlprofile/ | grep -oP "href=\"\K[^\"]*$(lsb_release -rs)[^\"]*\.deb")
+        echo "##vso[task.setvariable variable=packageName;isreadonly=true]$packageName"
+    ${{ if eq(parameters.os, 'almalinux8') }}:
+      script: |
+        export packageName=$(curl -s https://repo.radeon.com/rocm/rhel8/$(REPO_RADEON_VERSION)/main/ | grep -oP "hsa-amd-aqlprofile-[^\"]+\.rpm" | head -n1)
+        echo "##vso[task.setvariable variable=packageName;isreadonly=true]$packageName"
+- task: Bash@3
+  displayName: 'Download aqlprofile'
+  inputs:
+    targetType: inline
+    workingDirectory: '$(Pipeline.Workspace)'
+    ${{ if eq(parameters.os, 'ubuntu2204') }}:
+      script: wget -nv https://repo.radeon.com/rocm/apt/$(REPO_RADEON_VERSION)/pool/main/h/hsa-amd-aqlprofile/$(packageName)
+    ${{ if eq(parameters.os, 'almalinux8') }}:
+      script: wget -nv https://repo.radeon.com/rocm/rhel8/$(REPO_RADEON_VERSION)/main/$(packageName)
+- task: Bash@3
+  displayName: 'Extract aqlprofile'
+  inputs:
+    targetType: inline
+    workingDirectory: '$(Pipeline.Workspace)'
+    ${{ if eq(parameters.os, 'ubuntu2204') }}:
+      script: |
+        mkdir hsa-amd-aqlprofile
+        dpkg-deb -R $(packageName) hsa-amd-aqlprofile
+    ${{ if eq(parameters.os, 'almalinux8') }}:
+      script: |
+        mkdir hsa-amd-aqlprofile
+        sudo dnf -y install rpm-build cpio
+        rpm2cpio $(packageName) | (cd hsa-amd-aqlprofile && cpio -idmv)
+- task: Bash@3
+  displayName: 'Copy aqlprofile files'
  inputs:
    targetType: inline
-    workingDirectory: $(Agent.BuildDirectory)
    script: |
-      set -e
-      if [ "${{ parameters.os }}" = "ubuntu2204" ]; then
-        packageName=$(curl -s https://repo.radeon.com/rocm/apt/$(REPO_RADEON_VERSION)/pool/main/h/hsa-amd-aqlprofile/ | grep -oP "href=\"\K[^\"]*$(lsb_release -rs)[^\"]*\.deb") && \
-        wget -nv https://repo.radeon.com/rocm/apt/$(REPO_RADEON_VERSION)/pool/main/h/hsa-amd-aqlprofile/$packageName && \
-        mkdir -p hsa-amd-aqlprofile && \
-        dpkg-deb -R $packageName hsa-amd-aqlprofile
-      elif [ "${{ parameters.os }}" = "almalinux8" ]; then
-        sudo dnf -y install rpm-build cpio && \
-        packageName=$(curl -s https://repo.radeon.com/rocm/rhel8/$(REPO_RADEON_VERSION)/main/ | grep -oP "hsa-amd-aqlprofile-[^\"]+\.rpm" | head -n1) && \
-        wget -nv https://repo.radeon.com/rocm/rhel8/$(REPO_RADEON_VERSION)/main/$packageName && \
-        mkdir -p hsa-amd-aqlprofile && \
-        rpm2cpio $packageName | (cd hsa-amd-aqlprofile && cpio -idmv)
-      else
-        echo "Unsupported OS: ${{ parameters.os }}"
-        exit 1
-      fi && \
-      mkdir -p $(Agent.BuildDirectory)/rocm && \
-      cp -R hsa-amd-aqlprofile/opt/rocm-*/* $(Agent.BuildDirectory)/rocm && \
-      rm -rf hsa-amd-aqlprofile $packageName
+      mkdir -p $(Agent.BuildDirectory)/rocm
+      cp -R hsa-amd-aqlprofile/opt/rocm-*/* $(Agent.BuildDirectory)/rocm
+    workingDirectory: '$(Pipeline.Workspace)'
+- task: Bash@3
+  displayName: 'Clean up aqlprofile'
+  inputs:
+    targetType: inline
+    script: rm -rf hsa-amd-aqlprofile $(packageName)
+    workingDirectory: '$(Pipeline.Workspace)'
--- a/.azuredevops/templates/steps/dependencies-dnf.yml
+++ b/.azuredevops/templates/steps/dependencies-dnf.yml
@@ -54,13 +54,11 @@ parameters:
    libfftw3-dev: fftw-devel
    libfmt-dev: fmt-devel
    libgmp-dev: gmp-devel
-    liblapack-dev: lapack-devel
    liblzma-dev: xz-devel
    libmpfr-dev: mpfr-devel
    libmsgpack-dev: msgpack-devel
    libncurses5-dev: ncurses-devel
    libnuma-dev: numactl-devel
-    libopenblas-dev: openblas-devel
    libopenmpi-dev: openmpi-devel
    libpci-dev: libpciaccess-devel
    libssl-dev: openssl-devel
@@ -89,7 +87,6 @@ steps:
 - ${{ if eq(parameters.registerROCmPackages, true) }}:
  - task: Bash@3
    displayName: 'Register AMDGPU & ROCm repos (dnf)'
-    retryCountOnTaskFailure: 3
    inputs:
      targetType: inline
      script: |
@@ -110,13 +107,12 @@ steps:
        sudo dnf makecache
 - task: Bash@3
  displayName: 'Install base dnf packages'
-  retryCountOnTaskFailure: 3
  inputs:
    targetType: inline
    script: |
+      sudo dnf config-manager --set-enabled powertools
      # rpm fusion free repo for some dependencies
-      sudo dnf config-manager --set-enabled powertools && \
-      sudo dnf -y install https://download1.rpmfusion.org/free/el/rpmfusion-free-release-8.noarch.rpm && \
+      sudo dnf -y install https://download1.rpmfusion.org/free/el/rpmfusion-free-release-8.noarch.rpm
      sudo dnf -y install ${{ join(' ', parameters.basePackages) }}
 - task: Bash@3
  displayName: 'Check gcc environment'
@@ -130,7 +126,6 @@ steps:
      g++ -print-file-name=libstdc++.so
 - task: Bash@3
  displayName: 'Set python 3.11 as default'
-  retryCountOnTaskFailure: 3
  inputs:
    targetType: inline
    script: |
@@ -145,20 +140,18 @@ steps:
  - ${{ if eq(pkg, 'ninja-build') }}:
    - task: Bash@3
      displayName: 'Install ninja 1.11.1'
-      retryCountOnTaskFailure: 3
      inputs:
        targetType: inline
        script: |
-          sudo dnf -y install unzip && \
-          curl -LO https://github.com/ninja-build/ninja/releases/download/v1.11.1/ninja-linux.zip && \
-          unzip ninja-linux.zip && \
-          sudo mv ninja /usr/local/bin/ninja && \
-          sudo chmod +x /usr/local/bin/ninja && \
+          curl -LO https://github.com/ninja-build/ninja/releases/download/v1.11.1/ninja-linux.zip
+          sudo dnf -y install unzip
+          unzip ninja-linux.zip
+          sudo mv ninja /usr/local/bin/ninja
+          sudo chmod +x /usr/local/bin/ninja
          echo "##vso[task.prependpath]/usr/local/bin"
  - ${{ if ne(parameters.aptToDnfMap[pkg], '') }}:
    - task: Bash@3
      displayName: 'dnf install ${{ parameters.aptToDnfMap[pkg] }}'
-      retryCountOnTaskFailure: 3
      inputs:
        targetType: inline
        script: |
--- a/.azuredevops/templates/steps/dependencies-other.yml
+++ b/.azuredevops/templates/steps/dependencies-other.yml
@@ -27,7 +27,6 @@ steps:
 - ${{ if gt(length(parameters.pipModules), 0) }}:
  - task: Bash@3
    displayName: 'pip install  ...'
-    retryCountOnTaskFailure: 3
    inputs:
      targetType: inline
      script: python3 -m pip install -v --force-reinstall ${{ join(' ', parameters.pipModules) }}
--- a/.azuredevops/templates/steps/dependencies-rocm.yml
+++ b/.azuredevops/templates/steps/dependencies-rocm.yml
@@ -47,8 +47,8 @@ parameters:
      developBranch: aomp-dev
      hasGpuTarget: false
    clr:
-      pipelineId: 335
-      developBranch: develop
+      pipelineId: 145
+      developBranch: amd-staging
      hasGpuTarget: false
    composable_kernel:
      pipelineId: 86
@@ -59,8 +59,8 @@ parameters:
      developBranch: rocm
      hasGpuTarget: false
    HIP:
-      pipelineId: 335
-      developBranch: develop
+      pipelineId: 93
+      developBranch: amd-staging
      hasGpuTarget: false
    hip-tests:
      pipelineId: 233
@@ -203,8 +203,8 @@ parameters:
      developBranch: develop
      hasGpuTarget: true
    rocprofiler:
-      pipelineId: 329
-      developBranch: develop
+      pipelineId: 143
+      developBranch: amd-staging
      hasGpuTarget: true
    rocprofiler-compute:
      pipelineId: 257
--- a/.azuredevops/templates/steps/dependencies-vendor.yml
+++ b/.azuredevops/templates/steps/dependencies-vendor.yml
@@ -8,18 +8,15 @@ parameters:
  type: object
  default:
    boost: 250
-    fmtlib: 341
    grpc: 72
    gtest: 73
    half560: 68
    lapack: 69
-    spdlog: 340

 steps:
 - ${{ each dependency in parameters.dependencyList }}:
  - task: DownloadPipelineArtifact@2
    displayName: Download ${{ dependency }}
-    retryCountOnTaskFailure: 3
    inputs:
      project: ROCm-CI
      buildType: specific
--- a/.azuredevops/templates/steps/local-artifact-download.yml
+++ b/.azuredevops/templates/steps/local-artifact-download.yml
@@ -33,7 +33,6 @@ parameters:
 steps:
  - task: DownloadPipelineArtifact@2
    displayName: Download ${{ parameters.preTargetFilter}}*${{ parameters.os }}_${{ parameters.gpuTarget}}*${{ parameters.postTargetFilter}}
-    retryCountOnTaskFailure: 3
    inputs:
      ${{ if eq(parameters.buildType, 'specific') }}:
        buildType: specific
--- a/.azuredevops/templates/steps/miopen-get-ck-build.yml
+++ b/.azuredevops/templates/steps/miopen-get-ck-build.yml
@@ -7,7 +7,6 @@ steps:
 - task: Bash@3
  name: downloadCKBuild
  displayName: Download specific CK build
-  retryCountOnTaskFailure: 3
  env:
    CXX: $(Agent.BuildDirectory)/rocm/llvm/bin/amdclang++
    CC: $(Agent.BuildDirectory)/rocm/llvm/bin/amdclang
@@ -70,29 +69,20 @@ steps:

      RETRIES=0
      MAX_RETRIES=5
-      SUCCESS=false
-      while [ $RETRIES -lt $MAX_RETRIES ]; do
-        wget -nv $ARTIFACT_URL -O $(System.ArtifactsDirectory)/ck.zip && \
-          unzip $(System.ArtifactsDirectory)/ck.zip -d $(System.ArtifactsDirectory) && \
-          mkdir -p $(Agent.BuildDirectory)/rocm && \
-          tar -zxvf $(System.ArtifactsDirectory)/composable_kernel*/*.tar.gz -C $(Agent.BuildDirectory)/rocm && \
-          rm -r $(System.ArtifactsDirectory)/ck.zip $(System.ArtifactsDirectory)/composable_kernel*
-
-        if [ $? -eq 0 ]; then
-          SUCCESS=true
-          echo "Successfully downloaded CK."
-          break
-        else
-          RETRIES=$((RETRIES + 1))
-          echo "Failed to download CK on attempt $RETRIES/$MAX_RETRIES, retrying..."
-          sleep 1
+      until wget -nv $ARTIFACT_URL -O $(System.ArtifactsDirectory)/ck.zip; do
+        RETRIES=$((RETRIES+1))
+        if [[ $RETRIES -ge $MAX_RETRIES ]]; then
+          echo "Failed to download CK artifact after $MAX_RETRIES attempts."
+          exit 1
        fi
+        echo "Download failed, retrying ($RETRIES/$MAX_RETRIES)..."
+        sleep 5
      done

-      if [ "$SUCCESS" = false ]; then
-        echo "ERROR: failed to download CK after $MAX_RETRIES attempts."
-        exit 1
-      fi
+      unzip $(System.ArtifactsDirectory)/ck.zip -d $(System.ArtifactsDirectory)
+      mkdir -p $(Agent.BuildDirectory)/rocm
+      tar -zxvf $(System.ArtifactsDirectory)/composable_kernel*/*.tar.gz -C $(Agent.BuildDirectory)/rocm
+      rm -r $(System.ArtifactsDirectory)/ck.zip $(System.ArtifactsDirectory)/composable_kernel*

      if [[ $EXIT_CODE -ne 0 ]]; then
        BUILD_COMMIT=$(curl -s $AZ_API/build/builds/$CK_BUILD_ID | jq '.sourceVersion' | tr -d '"')
--- a/.azuredevops/variables-global.yml
+++ b/.azuredevops/variables-global.yml
@@ -28,13 +28,13 @@ variables:
 - name: GFX90A_TEST_POOL
  value: gfx90a_test_pool
 - name: LATEST_RELEASE_VERSION
-  value: 6.4.3
+  value: 6.4.2
 - name: REPO_RADEON_VERSION
-  value: 6.4.3
+  value: 6.4.2
 - name: NEXT_RELEASE_VERSION
  value: 7.0.0
 - name: LATEST_RELEASE_TAG
-  value: rocm-6.4.3
+  value: rocm-6.4.2
 - name: DOCKER_SKIP_GFX
  value: gfx90a
 - name: COMPOSABLE_KERNEL_PIPELINE_ID
--- a/.wordlist.txt
+++ b/.wordlist.txt
@@ -918,6 +918,7 @@ toolchain
 toolchains
 toolset
 toolsets
+torchtitan
 torchvision
 tqdm
 tracebacks
--- a/default.xml
+++ b/default.xml
@@ -1,7 +1,7 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <manifest>
    <remote name="rocm-org" fetch="https://github.com/ROCm/" />
-    <default revision="refs/tags/rocm-6.4.3"
+    <default revision="refs/tags/rocm-6.4.2"
     remote="rocm-org"
     sync-c="true"
     sync-j="4" />
--- a/docs/about/license.md
+++ b/docs/about/license.md
@@ -29,7 +29,6 @@ additional licenses. Please review individual repositories for more information.
 | [AMD SMI](https://github.com/ROCm/amdsmi) | [MIT](https://github.com/ROCm/amdsmi/blob/amd-staging/LICENSE) |
 | [aomp](https://github.com/ROCm/aomp/) | [Apache 2.0](https://github.com/ROCm/aomp/blob/aomp-dev/LICENSE) |
 | [aomp-extras](https://github.com/ROCm/aomp-extras/) | [MIT](https://github.com/ROCm/aomp-extras/blob/aomp-dev/LICENSE) |
-| [AQLprofile] | [MIT](https://github.com/ROCm/aqlprofile/blob/amd-staging/LICENSE) |
 | [Code Object Manager (Comgr)](https://github.com/ROCm/llvm-project/tree/amd-staging/amd/comgr) | [The University of Illinois/NCSA](https://github.com/ROCm/llvm-project/blob/amd-staging/amd/comgr/LICENSE.txt) |
 | [Composable Kernel](https://github.com/ROCm/composable_kernel) | [MIT](https://github.com/ROCm/composable_kernel/blob/develop/LICENSE) |
 | [half](https://github.com/ROCm/half/) | [MIT](https://github.com/ROCm/half/blob/rocm/LICENSE.txt) |
@@ -47,6 +46,7 @@ additional licenses. Please review individual repositories for more information.
 | [hipSPARSE](https://github.com/ROCm/hipSPARSE/) | [MIT](https://github.com/ROCm/hipSPARSE/blob/develop/LICENSE.md) |
 | [hipSPARSELt](https://github.com/ROCm/hipSPARSELt/) | [MIT](https://github.com/ROCm/hipSPARSELt/blob/develop/LICENSE.md) |
 | [hipTensor](https://github.com/ROCm/hipTensor) | [MIT](https://github.com/ROCm/hipTensor/blob/develop/LICENSE) |
+| hsa-amd-aqlprofile | [AMD Software EULA](https://www.amd.com/en/legal/eula/amd-software-eula.html) |
 | [llvm-project](https://github.com/ROCm/llvm-project/) | [Apache](https://github.com/ROCm/llvm-project/blob/amd-staging/LICENSE.TXT) |
 | [llvm-project/flang](https://github.com/ROCm/llvm-project/tree/amd-staging/flang) | [Apache 2.0](https://github.com/ROCm/llvm-project/blob/amd-staging/flang/LICENSE.TXT) |
 | [MIGraphX](https://github.com/ROCm/AMDMIGraphX/) | [MIT](https://github.com/ROCm/AMDMIGraphX/blob/develop/LICENSE) |
@@ -132,10 +132,12 @@ companies.
 ### Package licensing

 :::{attention}
-ROCprof Trace Decoder and AOCC CPU optimizations are provided in binary form, subject to the license agreement enclosed on [GitHub](https://github.com/ROCm/rocprof-trace-decoder/blob/amd-mainline/LICENSE) for ROCprof Trace Decoder, and [Developer Central](https://www.amd.com/en/developer/aocc.html) for AOCC. By using, installing,
-copying or distributing ROCprof Trace Decoder or AOCC CPU Optimizations, you agree to
+AQL Profiler and AOCC CPU optimization are both provided in binary form, each
+subject to the license agreement enclosed in the directory for the binary available
+in `/opt/rocm/share/doc/hsa-amd-aqlprofile/EULA`. By using, installing,
+copying or distributing AQL Profiler and/or AOCC CPU Optimizations, you agree to
 the terms and conditions of this license agreement. If you do not agree to the
-terms of this agreement, do not install, copy or use ROCprof Trace Decoder or the
+terms of this agreement, do not install, copy or use the AQL Profiler and/or the
 AOCC CPU Optimizations.
 :::

--- a/docs/conf.py
+++ b/docs/conf.py
@@ -9,21 +9,17 @@ import shutil
 import sys
 from pathlib import Path

-gh_release_path = os.path.join("..", "RELEASE.md")
-gh_changelog_path = os.path.join("..", "CHANGELOG.md")
-sphinx_release_path = os.path.join("about", "release-notes.md")
-sphinx_changelog_path = os.path.join("release", "changelog.md")
-shutil.copy2(gh_release_path, sphinx_release_path)
-shutil.copy2(gh_changelog_path, sphinx_changelog_path)
+shutil.copy2("../RELEASE.md", "./about/release-notes.md")
+shutil.copy2("../CHANGELOG.md", "./release/changelog.md")

 # Mark the consolidated changelog as orphan to prevent Sphinx from warning about missing toctree entries
-with open(sphinx_changelog_path, "r+", encoding="utf-8") as file:
+with open("./release/changelog.md", "r+") as file:
    content = file.read()
    file.seek(0)
    file.write(":orphan:\n" + content)

 # Replace GitHub-style [!ADMONITION]s with Sphinx-compatible ```{admonition} blocks
-with open(sphinx_changelog_path, "r", encoding="utf-8") as file:
+with open("./release/changelog.md", "r") as file:
    lines = file.readlines()

    modified_lines = []
@@ -61,14 +57,11 @@ with open(sphinx_changelog_path, "r", encoding="utf-8") as file:

    file.close()

-    with open(sphinx_changelog_path, "w", encoding="utf-8") as file:
+    with open("./release/changelog.md", 'w') as file:
        file.writelines(modified_lines)

-matrix_path = os.path.join("compatibility", "compatibility-matrix-historical-6.0.csv")
-rtd_path = os.path.join("..", "_readthedocs", "html", "downloads")
-if not os.path.exists(rtd_path):
-    os.makedirs(rtd_path)
-shutil.copy2(matrix_path, rtd_path)
+os.system("mkdir -p ../_readthedocs/html/downloads")
+os.system("cp compatibility/compatibility-matrix-historical-6.0.csv ../_readthedocs/html/downloads/compatibility-matrix-historical-6.0.csv")

 latex_engine = "xelatex"
 latex_elements = {
@@ -124,11 +117,15 @@ article_pages = [
    {"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.3", "os": ["linux"]},
    {"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.4", "os": ["linux"]},
    {"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.5", "os": ["linux"]},
+    {"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.6", "os": ["linux"]},
+    {"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-primus-migration-guide", "os": ["linux"]},
+    {"file": "how-to/rocm-for-ai/training/benchmark-docker/primus-megatron", "os": ["linux"]},
    {"file": "how-to/rocm-for-ai/training/benchmark-docker/pytorch-training", "os": ["linux"]},
    {"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-history", "os": ["linux"]},
    {"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.3", "os": ["linux"]},
    {"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.4", "os": ["linux"]},
    {"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.5", "os": ["linux"]},
+    {"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.6", "os": ["linux"]},
    {"file": "how-to/rocm-for-ai/training/benchmark-docker/jax-maxtext", "os": ["linux"]},
    {"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-history", "os": ["linux"]},
    {"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-v25.4", "os": ["linux"]},
--- a/docs/contribute/building.md
+++ b/docs/contribute/building.md
@@ -28,31 +28,13 @@ See the [Python requirements file](https://github.com/ROCm/ROCm/blob/develop/doc

 Use the Python Virtual Environment (`venv`) and run the following commands from the project root:

-::::{tab-set}
-:::{tab-item} Linux and WSL
-:sync: linux
-
 ```sh
 python3 -mvenv .venv

-.venv/bin/python -m pip install -r docs/sphinx/requirements.txt
-.venv/bin/python -m sphinx -T -E -b html -d _build/doctrees -D language=en docs _build/html
+.venv/bin/python     -m pip install -r docs/sphinx/requirements.txt
+.venv/bin/python     -m sphinx -T -E -b html -d _build/doctrees -D language=en docs _build/html
 ```

-:::
-:::{tab-item} Windows
-:sync: windows
-
-```powershell
-python -mvenv .venv
-
-.venv\Scripts\python.exe -m pip install -r docs/sphinx/requirements.txt
-.venv\Scripts\python.exe -m sphinx -T -E -b html -d _build/doctrees -D language=en docs _build/html
-```
-
-:::
-::::
-
 Navigate to `_build/html/index.html` and open this file in a web browser.

 ## Visual Studio Code
--- a/docs/data/how-to/rocm-for-ai/training/previous-versions/pytorch-training-v25.6-benchmark-models.yaml
+++ b/docs/data/how-to/rocm-for-ai/training/previous-versions/pytorch-training-v25.6-benchmark-models.yaml
@@ -0,0 +1,120 @@
+unified_docker:
+  latest:
+    pull_tag: rocm/pytorch-training:v25.6
+    docker_hub_url: https://hub.docker.com/r/rocm/pytorch-training/tags
+    rocm_version: 6.4.1
+    pytorch_version: 2.8.0a0+git7d205b2
+    python_version: 3.10.17
+    transformer_engine_version: 1.14.0+2f85f5f2
+    flash_attention_version: 3.0.0.post1
+    hipblaslt_version: 0.15.0-8c6919d
+    triton_version: 3.3.0
+model_groups:
+  - group: Pre-training
+    tag: pre-training
+    models:
+    - model: Llama 3.1 8B
+      mad_tag: pyt_train_llama-3.1-8b
+      model_repo: Llama-3.1-8B
+      url: https://huggingface.co/meta-llama/Llama-3.1-8B
+      precision: BF16
+      training_modes: [pretrain]
+    - model: Llama 3.1 70B
+      mad_tag: pyt_train_llama-3.1-70b
+      model_repo: Llama-3.1-70B
+      url: https://huggingface.co/meta-llama/Llama-3.1-70B-Instruct
+      precision: BF16
+      training_modes: [pretrain]
+    - model: FLUX.1-dev
+      mad_tag: pyt_train_flux
+      model_repo: Flux
+      url: https://huggingface.co/black-forest-labs/FLUX.1-dev
+      precision: BF16
+      training_modes: [pretrain]
+  - group: Fine-tuning
+    tag: fine-tuning
+    models:
+    - model: Llama 4 Scout 17B-16E
+      mad_tag: pyt_train_llama-4-scout-17b-16e
+      model_repo: Llama-4-17B_16E
+      url: https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora]
+    - model: Llama 3.3 70B
+      mad_tag: pyt_train_llama-3.3-70b
+      model_repo: Llama-3.3-70B
+      url: https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora, finetune_qlora]
+    - model: Llama 3.2 1B
+      mad_tag: pyt_train_llama-3.2-1b
+      model_repo: Llama-3.2-1B
+      url: https://huggingface.co/meta-llama/Llama-3.2-1B
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora]
+    - model: Llama 3.2 3B
+      mad_tag: pyt_train_llama-3.2-3b
+      model_repo: Llama-3.2-3B
+      url: https://huggingface.co/meta-llama/Llama-3.2-3B
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora]
+    - model: Llama 3.2 Vision 11B
+      mad_tag: pyt_train_llama-3.2-vision-11b
+      model_repo: Llama-3.2-Vision-11B
+      url: https://huggingface.co/meta-llama/Llama-3.2-11B-Vision
+      precision: BF16
+      training_modes: [finetune_fw]
+    - model: Llama 3.2 Vision 90B
+      mad_tag: pyt_train_llama-3.2-vision-90b
+      model_repo: Llama-3.2-Vision-90B
+      url: https://huggingface.co/meta-llama/Llama-3.2-90B-Vision
+      precision: BF16
+      training_modes: [finetune_fw]
+    - model: Llama 3.1 8B
+      mad_tag: pyt_train_llama-3.1-8b
+      model_repo: Llama-3.1-8B
+      url: https://huggingface.co/meta-llama/Llama-3.1-8B
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora]
+    - model: Llama 3.1 70B
+      mad_tag: pyt_train_llama-3.1-70b
+      model_repo: Llama-3.1-70B
+      url: https://huggingface.co/meta-llama/Llama-3.1-70B
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora, finetune_qlora]
+    - model: Llama 3.1 405B
+      mad_tag: pyt_train_llama-3.1-405b
+      model_repo: Llama-3.1-405B
+      url: https://huggingface.co/meta-llama/Llama-3.1-405B
+      precision: BF16
+      training_modes: [finetune_qlora, HF_finetune_lora]
+    - model: Llama 3 8B
+      mad_tag: pyt_train_llama-3-8b
+      model_repo: Llama-3-8B
+      url: https://huggingface.co/meta-llama/Meta-Llama-3-8B
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora]
+    - model: Llama 3 70B
+      mad_tag: pyt_train_llama-3-70b
+      model_repo: Llama-3-70B
+      url: https://huggingface.co/meta-llama/Meta-Llama-3-70B
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora]
+    - model: Llama 2 7B
+      mad_tag: pyt_train_llama-2-7b
+      model_repo: Llama-2-7B
+      url: https://github.com/meta-llama/llama-models/tree/main/models/llama2
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora, finetune_qlora]
+    - model: Llama 2 13B
+      mad_tag: pyt_train_llama-2-13b
+      model_repo: Llama-2-13B
+      url: https://github.com/meta-llama/llama-models/tree/main/models/llama2
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora]
+    - model: Llama 2 70B
+      mad_tag: pyt_train_llama-2-70b
+      model_repo: Llama-2-70B
+      url: https://github.com/meta-llama/llama-models/tree/main/models/llama2
+      precision: BF16
+      training_modes: [finetune_lora, finetune_qlora, HF_finetune_lora]
--- a/docs/data/how-to/rocm-for-ai/training/pytorch-training-benchmark-models.yaml
+++ b/docs/data/how-to/rocm-for-ai/training/pytorch-training-benchmark-models.yaml
@@ -1,38 +1,17 @@
-unified_docker:
-  latest:
-    pull_tag: rocm/pytorch-training:v25.6
-    docker_hub_url: https://hub.docker.com/r/rocm/pytorch-training/tags
-    rocm_version: 6.4.1
-    pytorch_version: 2.8.0a0+git7d205b2
-    python_version: 3.10.17
-    transformer_engine_version: 1.14.0+2f85f5f2
-    flash_attention_version: 3.0.0.post1
-    hipblaslt_version: 0.15.0-8c6919d
-    triton_version: 3.3.0
+dockers:
+  - pull_tag: rocm/pytorch-training:v25.7
+    docker_hub_url: https://hub.docker.com/layers/rocm/pytorch-training/v25.7/images/sha256-cc6fd840ab89cb81d926fc29eca6d075aee9875a55a522675a4b9231c9a0a712
+    components:
+      ROCm: 6.4.2
+      PyTorch: 2.8.0a0+gitd06a406
+      Python: 3.10.18
+      Transformer Engine: 2.2.0.dev0+94e53dd8
+      Flash Attention: 3.0.0.post1
+      hipBLASLt: 1.1.0-4b9a52edfc
+      Triton: 3.3.0
 model_groups:
-  - group: Pre-training
-    tag: pre-training
-    models:
-    - model: Llama 3.1 8B
-      mad_tag: pyt_train_llama-3.1-8b
-      model_repo: Llama-3.1-8B
-      url: https://huggingface.co/meta-llama/Llama-3.1-8B
-      precision: BF16
-      training_modes: [pretrain]
-    - model: Llama 3.1 70B
-      mad_tag: pyt_train_llama-3.1-70b
-      model_repo: Llama-3.1-70B
-      url: https://huggingface.co/meta-llama/Llama-3.1-70B-Instruct
-      precision: BF16
-      training_modes: [pretrain]
-    - model: FLUX.1-dev
-      mad_tag: pyt_train_flux
-      model_repo: Flux
-      url: https://huggingface.co/black-forest-labs/FLUX.1-dev
-      precision: BF16
-      training_modes: [pretrain]
-  - group: Fine-tuning
-    tag: fine-tuning
+  - group: Meta Llama
+    tag: llama
    models:
    - model: Llama 4 Scout 17B-16E
      mad_tag: pyt_train_llama-4-scout-17b-16e
@@ -75,19 +54,19 @@ model_groups:
      model_repo: Llama-3.1-8B
      url: https://huggingface.co/meta-llama/Llama-3.1-8B
      precision: BF16
-      training_modes: [finetune_fw, finetune_lora]
+      training_modes: [pretrain, finetune_fw, finetune_lora, HF_pretrain]
    - model: Llama 3.1 70B
      mad_tag: pyt_train_llama-3.1-70b
      model_repo: Llama-3.1-70B
-      url: https://huggingface.co/meta-llama/Llama-3.1-70B
+      url: https://huggingface.co/meta-llama/Llama-3.1-70B-Instruct
      precision: BF16
-      training_modes: [finetune_fw, finetune_lora, finetune_qlora]
+      training_modes: [pretrain, finetune_fw, finetune_lora]
    - model: Llama 3.1 405B
      mad_tag: pyt_train_llama-3.1-405b
      model_repo: Llama-3.1-405B
      url: https://huggingface.co/meta-llama/Llama-3.1-405B
      precision: BF16
-      training_modes: [finetune_qlora, HF_finetune_lora]
+      training_modes: [finetune_qlora]
    - model: Llama 3 8B
      mad_tag: pyt_train_llama-3-8b
      model_repo: Llama-3-8B
@@ -117,4 +96,67 @@ model_groups:
      model_repo: Llama-2-70B
      url: https://github.com/meta-llama/llama-models/tree/main/models/llama2
      precision: BF16
-      training_modes: [finetune_lora, finetune_qlora, HF_finetune_lora]
+      training_modes: [finetune_lora, finetune_qlora]
+  - group: OpenAI
+    tag: openai
+    models:
+    - model: GPT OSS 20B
+      mad_tag: pyt_train_gpt_oss_20b
+      model_repo: GPT-OSS-20B
+      url: https://huggingface.co/openai/gpt-oss-20b
+      precision: BF16
+      training_modes: [HF_finetune_lora]
+    - model: GPT OSS 120B
+      mad_tag: pyt_train_gpt_oss_120b
+      model_repo: GPT-OSS-120B
+      url: https://huggingface.co/openai/gpt-oss-120b
+      precision: BF16
+      training_modes: [HF_finetune_lora]
+  - group: Qwen
+    tag: qwen
+    models:
+    - model: Qwen 3 8B
+      mad_tag: pyt_train_qwen3-8b
+      model_repo: Qwen3-8B
+      url: https://huggingface.co/Qwen/Qwen3-8B
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora]
+    - model: Qwen 3 32B
+      mad_tag: pyt_train_qwen3-32b
+      model_repo: Qwen3-32
+      url: https://huggingface.co/Qwen/Qwen3-32B
+      precision: BF16
+      training_modes: [finetune_lora]
+    - model: Qwen 2.5 32B
+      mad_tag: pyt_train_qwen2.5-32b
+      model_repo: Qwen2.5-32B
+      url: https://huggingface.co/Qwen/Qwen2.5-32B
+      precision: BF16
+      training_modes: [finetune_lora]
+    - model: Qwen 2.5 72B
+      mad_tag: pyt_train_qwen2.5-72b
+      model_repo: Qwen2.5-72B
+      url: https://huggingface.co/Qwen/Qwen2.5-72B
+      precision: BF16
+      training_modes: [finetune_lora]
+    - model: Qwen 2 1.5B
+      mad_tag: pyt_train_qwen2-1.5b
+      model_repo: Qwen2-1.5B
+      url: https://huggingface.co/Qwen/Qwen2-1.5B
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora]
+    - model: Qwen 2 7B
+      mad_tag: pyt_train_qwen2-7b
+      model_repo: Qwen2-7B
+      url: https://huggingface.co/Qwen/Qwen2-7B
+      precision: BF16
+      training_modes: [finetune_fw, finetune_lora]
+  - group: Flux
+    tag: flux
+    models:
+    - model: FLUX.1-dev
+      mad_tag: pyt_train_flux
+      model_repo: Flux
+      url: https://huggingface.co/black-forest-labs/FLUX.1-dev
+      precision: BF16
+      training_modes: [pretrain]
--- a/docs/how-to/deep-learning-rocm.rst
+++ b/docs/how-to/deep-learning-rocm.rst
@@ -23,93 +23,92 @@ The table below summarizes information about ROCm-enabled deep learning framewor
      - Installation options
      - GitHub

-    * - `PyTorch <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/pytorch-compatibility.html>`_
+    * - `PyTorch <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/pytorch-compatibility.html>`__
      - .. raw:: html
-         
+
          <a href="https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/pytorch-install.html"><i class="fas fa-link fa-lg"></i></a>
      - 
-        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/pytorch-install.html#using-a-docker-image-with-pytorch-pre-installed>`_ 
-        - `Wheels package <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/pytorch-install.html#using-a-wheels-package>`_
-        - `ROCm Base Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/pytorch-install.html#using-the-pytorch-rocm-base-docker-image>`_ 
-        - `Upstream Docker file <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/pytorch-install.html#using-the-pytorch-upstream-dockerfile>`_
+        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/pytorch-install.html#using-a-docker-image-with-pytorch-pre-installed>`__
+        - `Wheels package <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/pytorch-install.html#using-a-wheels-package>`__
+        - `ROCm Base Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/pytorch-install.html#using-the-pytorch-rocm-base-docker-image>`__
+        - `Upstream Docker file <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/pytorch-install.html#using-the-pytorch-upstream-dockerfile>`__
      - .. raw:: html
-         
+
          <a href="https://github.com/ROCm/pytorch"><i class="fab fa-github fa-lg"></i></a>
-   
-    * - `TensorFlow <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/tensorflow-compatibility.html>`_
+
+    * - `TensorFlow <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/tensorflow-compatibility.html>`__
      - .. raw:: html
-         
+
          <a href="https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/tensorflow-install.html"><i class="fas fa-link fa-lg"></i></a>
      - 
-        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/tensorflow-install.html#using-a-docker-image-with-tensorflow-pre-installed>`_
-        - `Wheels package <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/tensorflow-install.html#using-a-wheels-package>`_
+        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/tensorflow-install.html#using-a-docker-image-with-tensorflow-pre-installed>`__
+        - `Wheels package <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/tensorflow-install.html#using-a-wheels-package>`__

      - .. raw:: html
-         
+
          <a href="https://github.com/ROCm/tensorflow-upstream"><i class="fab fa-github fa-lg"></i></a> 

-    * - `JAX <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/jax-compatibility.html>`_
+    * - `JAX <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/jax-compatibility.html>`__
      - .. raw:: html
-         
+
          <a href="https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/jax-install.html"><i class="fas fa-link fa-lg"></i></a>
      - 
-        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/jax-install.html#using-a-prebuilt-docker-image>`_
+        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/jax-install.html#using-a-prebuilt-docker-image>`__
      - .. raw:: html
-         
+
          <a href="https://github.com/ROCm/jax"><i class="fab fa-github fa-lg"></i></a>
-   
-    * - `verl <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/verl-compatibility.html>`_
+
+    * - `verl <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/verl-compatibility.html>`__
      - .. raw:: html
-         
+
          <a href="https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/verl-install.html"><i class="fas fa-link fa-lg"></i></a>
      - 
-        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/verl-install.html#use-a-prebuilt-docker-image-with-verl-pre-installed>`_
+        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/verl-install.html#use-a-prebuilt-docker-image-with-verl-pre-installed>`__
      - .. raw:: html
-         
+
          <a href="https://github.com/ROCm/verl"><i class="fab fa-github fa-lg"></i></a>

-    * - `Stanford Megatron-LM <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/stanford-megatron-lm-compatibility.html>`_
+    * - `Stanford Megatron-LM <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/stanford-megatron-lm-compatibility.html>`__
      - .. raw:: html
-         
+
          <a href="https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/stanford-megatron-lm-install.html"><i class="fas fa-link fa-lg"></i></a>
      - 
-        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/stanford-megatron-lm-install.html#use-a-prebuilt-docker-image-with-stanford-megatron-lm-pre-installed>`_
+        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/stanford-megatron-lm-install.html#use-a-prebuilt-docker-image-with-stanford-megatron-lm-pre-installed>`__
      - .. raw:: html
-         
+
          <a href="https://github.com/ROCm/Stanford-Megatron-LM"><i class="fab fa-github fa-lg"></i></a>
-   
-    * - `DGL <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/dgl-compatibility.html>`_
+
+    * - `DGL <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/dgl-compatibility.html>`__
      - .. raw:: html
-         
+
          <a href="https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/dgl-install.html"><i class="fas fa-link fa-lg"></i></a>
      - 
-        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/dgl-install.html#use-a-prebuilt-docker-image-with-dgl-pre-installed>`_
+        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/dgl-install.html#use-a-prebuilt-docker-image-with-dgl-pre-installed>`__
      - .. raw:: html
-         
+
          <a href="https://github.com/ROCm/dgl"><i class="fab fa-github fa-lg"></i></a> 

-    * - `Megablocks <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/megablocks-compatibility.html>`_
+    * - `Megablocks <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/megablocks-compatibility.html>`__
      - .. raw:: html
-         
+
          <a href="https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/megablocks-install.html"><i class="fas fa-link fa-lg"></i></a>
      - 
-        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/megablocks-install.html#using-a-prebuilt-docker-image-with-megablocks-pre-installed>`_
+        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/megablocks-install.html#using-a-prebuilt-docker-image-with-megablocks-pre-installed>`__
      - .. raw:: html
-         
+
          <a href="https://github.com/ROCm/megablocks"><i class="fab fa-github fa-lg"></i></a>
-   
-    * - `Taichi <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/taichi-compatibility.html>`_
+
+    * - `Taichi <https://rocm.docs.amd.com/en/latest/compatibility/ml-compatibility/taichi-compatibility.html>`__
      - .. raw:: html
-         
+
          <a href="https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/taichi-install.html"><i class="fas fa-link fa-lg"></i></a>
      - 
-        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/taichi-install.html#use-a-prebuilt-docker-image-with-taichi-pre-installed>`_ 
-        - `Wheels package <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/taichi-install.html#use-a-wheels-package>`_
+        - `Docker image <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/taichi-install.html#use-a-prebuilt-docker-image-with-taichi-pre-installed>`__
+        - `Wheels package <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/taichi-install.html#use-a-wheels-package>`__

      - .. raw:: html
-         
-          <a href="https://github.com/ROCm/taichi"><i class="fab fa-github fa-lg"></i></a>      

+          <a href="https://github.com/ROCm/taichi"><i class="fab fa-github fa-lg"></i></a>

 Learn how to use your ROCm deep learning environment for training, fine-tuning, inference, and performance optimization
 through the following guides.
@@ -124,10 +123,3 @@ through the following guides.

 * :doc:`Use ROCm for AI inference optimization <rocm-for-ai/inference-optimization/index>`

-
-
-
-
-
-
-
--- a/docs/how-to/rocm-for-ai/inference-optimization/workload.rst
+++ b/docs/how-to/rocm-for-ai/inference-optimization/workload.rst
@@ -939,7 +939,7 @@ hipBLASLt benchmarking
 The GEMM library
 `hipBLASLt <https://rocm.docs.amd.com/projects/hipBLASLt/en/latest/index.html>`_
 provides a benchmark tool for its supported operations. Refer to the
-`documentation <https://github.com/ROCm/hipBLASLt/blob/develop/clients/benchmarks/README.md>`_
+`documentation <https://github.com/ROCm/hipBLASLt/blob/develop/clients/bench/README.md>`_
 for details.

 * Example 1: Benchmark mix fp8 GEMM
--- a/docs/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.9.1-20250715.rst
+++ b/docs/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.9.1-20250715.rst
@@ -46,7 +46,7 @@ vLLM inference performance testing
        - {{ unified_docker.hipblaslt_version }}

 With this Docker image, you can quickly test the :ref:`expected
-inference performance numbers <vllm-benchmark-performance-measurements>` for
+inference performance numbers <vllm-benchmark-performance-measurements-715>` for
 MI300X series accelerators.

 What's new
@@ -219,7 +219,7 @@ system's configuration.
            ``container_ci-{{model.mad_tag}}``. The latency and throughput reports of the
            model are collected in the following path: ``~/MAD/reports_{{model.precision}}/``.

-            Although the :ref:`available models <vllm-benchmark-available-models>` are preconfigured
+            Although the :ref:`available models <vllm-benchmark-available-models-715>` are preconfigured
            to collect latency and throughput performance data, you can also change the benchmarking
            parameters. See the standalone benchmarking tab for more information.

--- a/docs/how-to/rocm-for-ai/inference/benchmark-docker/vllm.rst
+++ b/docs/how-to/rocm-for-ai/inference/benchmark-docker/vllm.rst
@@ -39,7 +39,7 @@ vLLM inference performance testing
        - {{ unified_docker.hipblaslt_version }}

 With this Docker image, you can quickly test the :ref:`expected
-inference performance numbers <vllm-benchmark-performance-measurements>` for
+inference performance numbers <vllm-benchmark-performance-measurements-812>` for
 MI300X series accelerators.

 What's new
@@ -208,7 +208,7 @@ system's configuration.
            and ``{{ model.mad_tag }}_serving.csv``.

            Although the :ref:`available models
-            <vllm-benchmark-available-models>` are preconfigured to collect
+            <vllm-benchmark-available-models-812>` are preconfigured to collect
            offline throughput and online serving performance data, you can
            also change the benchmarking parameters. See the standalone
            benchmarking tab for more information.
--- a/docs/how-to/rocm-for-ai/install.rst
+++ b/docs/how-to/rocm-for-ai/install.rst
@@ -22,9 +22,9 @@ If you’re new to ROCm, refer to the :doc:`ROCm quick start install guide for L
 <rocm-install-on-linux:install/quick-start>`.

 If you’re using a Radeon GPU for graphics-accelerated applications, refer to the
-`Radeon installation instructions <https://rocm.docs.amd.com/projects/radeon/en/docs-6.1.3/docs/install/native_linux/install-radeon.html>`_.
+`Radeon installation instructions <https://rocm.docs.amd.com/projects/radeon/en/latest/docs/install/native_linux/howto_native_linux.html>`_.

-You can install ROCm on :ref:`compatible systems <rocm-install-on-linux:reference/system-requirements>` via your Linux
+You can install ROCm on :doc:`compatible systems <rocm-install-on-linux:reference/system-requirements>` via your Linux
 distribution's package manager. See the following documentation resources to get started:

 * :doc:`ROCm installation overview <rocm-install-on-linux:install/install-overview>`
--- a/docs/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v24.12-dev.rst
+++ b/docs/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v24.12-dev.rst
@@ -18,7 +18,7 @@ Training a model with ROCm Megatron-LM
 The ROCm Megatron-LM framework is a specialized fork of the robust Megatron-LM, designed to
 enable efficient training of large-scale language models on AMD GPUs. By leveraging AMD Instinct™ MI300X
 accelerators, AMD Megatron-LM delivers enhanced scalability, performance, and resource utilization for AI
-workloads. It is purpose-built to :ref:`support models <amd-megatron-lm-model-support>`
+workloads. It is purpose-built to :ref:`support models <amd-megatron-lm-model-support-24-12>`
 like Meta's Llama 2, Llama 3, and Llama 3.1, enabling developers to train next-generation AI models with greater
 efficiency. See the GitHub repository at `<https://github.com/ROCm/Megatron-LM>`__.

@@ -67,7 +67,7 @@ Megatron-LM provides the following key features to train large language models e

 - Pre-training

-.. _amd-megatron-lm-model-support:
+.. _amd-megatron-lm-model-support-24-12:

 The following models are pre-optimized for performance on the AMD Instinct MI300X accelerator.

--- a/docs/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.3.rst
+++ b/docs/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.3.rst
@@ -67,7 +67,7 @@ Megatron-LM provides the following key features to train large language models e

 - Pre-training

-.. _amd-megatron-lm-model-support:
+.. _amd-megatron-lm-model-support-25-3:

 The following models are pre-optimized for performance on the AMD Instinct MI300X accelerator.

@@ -278,7 +278,7 @@ handle a variety of input sequences, including unseen words or domain-specific t
   .. tab-item:: Llama
      :sync: llama

-      To train any of the Llama 2 models that :ref:`this Docker image supports <amd-megatron-lm-model-support>`, use the ``Llama2Tokenizer``.
+      To train any of the Llama 2 models that :ref:`this Docker image supports <amd-megatron-lm-model-support-25-3>`, use the ``Llama2Tokenizer``.

      To train any of Llama 3 and Llama 3.1 models that this Docker image supports, use the ``HuggingFaceTokenizer``.
      Set the Hugging Face model link in the ``TOKENIZER_MODEL`` variable.
@@ -292,7 +292,7 @@ handle a variety of input sequences, including unseen words or domain-specific t
   .. tab-item:: DeepSeek V2
      :sync: deepseek

-      To train any of the DeepSeek V2 models that :ref:`this Docker image supports <amd-megatron-lm-model-support>`, use the ``DeepSeekV2Tokenizer``.
+      To train any of the DeepSeek V2 models that :ref:`this Docker image supports <amd-megatron-lm-model-support-25-3>`, use the ``DeepSeekV2Tokenizer``.

 Multi-node training
 ^^^^^^^^^^^^^^^^^^^
--- a/docs/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.4.rst
+++ b/docs/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.4.rst
@@ -67,7 +67,7 @@ Megatron-LM provides the following key features to train large language models e

 - Pre-training

-.. _amd-megatron-lm-model-support:
+.. _amd-megatron-lm-model-support-25-4:

 The following models are pre-optimized for performance on AMD Instinct MI300X series accelerators.

@@ -291,7 +291,7 @@ or ``${DATA_DIR}/tokenizer_llama2``.
   .. tab-item:: Llama
      :sync: llama

-      To train any of the Llama 2 models that :ref:`this Docker image supports <amd-megatron-lm-model-support>`, use the ``Llama2Tokenizer``
+      To train any of the Llama 2 models that :ref:`this Docker image supports <amd-megatron-lm-model-support-25-4>`, use the ``Llama2Tokenizer``
      or the default ``HuggingFaceTokenizer``.

      To train any of Llama 3 and Llama 3.1 models that this Docker image supports, use the ``HuggingFaceTokenizer``.
@@ -320,7 +320,7 @@ or ``${DATA_DIR}/tokenizer_llama2``.
   .. tab-item:: DeepSeek V2
      :sync: deepseek

-      To train any of the DeepSeek V2 models that :ref:`this Docker image supports <amd-megatron-lm-model-support>`, use the ``DeepSeekV2Tokenizer``.
+      To train any of the DeepSeek V2 models that :ref:`this Docker image supports <amd-megatron-lm-model-support-25-4>`, use the ``DeepSeekV2Tokenizer``.

 Multi-node training
 ^^^^^^^^^^^^^^^^^^^
--- a/docs/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-history.rst
+++ b/docs/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-history.rst
@@ -16,12 +16,20 @@ previous releases of the ``ROCm/pytorch-training`` Docker image on `Docker Hub <
     - Components
     - Resources

+   * - v25.7
+     - 
+       * ROCm 6.4.2
+       * PyTorch 2.8.0a0+gitd06a406
+     - 
+       * :doc:`Documentation <../pytorch-training>`
+       * `Docker Hub <https://hub.docker.com/layers/rocm/pytorch-training/v25.7/images/sha256-cc6fd840ab89cb81d926fc29eca6d075aee9875a55a522675a4b9231c9a0a712>`__
+
   * - v25.6
     - 
       * ROCm 6.3.4
       * PyTorch 2.8.0a0+git7d205b2
     - 
-       * :doc:`Documentation <../pytorch-training>`
+       * :doc:`Documentation <pytorch-training-v25.6>`
       * `Docker Hub <https://hub.docker.com/layers/rocm/pytorch-training/v25.6/images/sha256-a4cea3c493a4a03d199a3e81960ac071d79a4a7a391aa9866add3b30a7842661>`__

   * - v25.5
--- a/docs/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.5.rst
+++ b/docs/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.5.rst
@@ -437,3 +437,8 @@ Once the setup is complete, choose between two options to start benchmarking:

           ./pytorch_benchmark_report.sh -t HF_finetune_lora -p BF16 -m Llama-2-70B

+Previous versions
+=================
+
+See :doc:`pytorch-training-history` to find documentation for previous releases
+of the ``ROCm/pytorch-training`` Docker image.
--- a/docs/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.6.rst
+++ b/docs/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.6.rst
@@ -0,0 +1,456 @@
+:orphan:
+
+.. meta::
+   :description: How to train a model using PyTorch for ROCm.
+   :keywords: ROCm, AI, LLM, train, PyTorch, torch, Llama, flux, tutorial, docker
+
+**************************************
+Training a model with PyTorch for ROCm
+**************************************
+
+.. caution::
+
+   This documentation does not reflect the latest version of ROCm vLLM
+   performance benchmark documentation. See :doc:`../pytorch-training` for the latest version.
+
+PyTorch is an open-source machine learning framework that is widely used for
+model training with GPU-optimized components for transformer-based models.
+
+The `PyTorch for ROCm training Docker <https://hub.docker.com/layers/rocm/pytorch-training/v25.6/images/sha256-a4cea3c493a4a03d199a3e81960ac071d79a4a7a391aa9866add3b30a7842661>`_
+(``rocm/pytorch-training:v25.6``) image provides a prebuilt optimized environment for fine-tuning and pretraining a
+model on AMD Instinct MI325X and MI300X accelerators. It includes the following software components to accelerate
+training workloads:
+
+--------------------------+--------------------------------+
+| Software component       | Version                        |
+==========================+================================+
+| ROCm                     | 6.3.4                          |
+--------------------------+--------------------------------+
+| PyTorch                  | 2.8.0a0+git7d205b2             |
+--------------------------+--------------------------------+
+| Python                   | 3.10.17                        |
+--------------------------+--------------------------------+
+| Transformer Engine       | 1.14.0+2f85f5f2                |
+--------------------------+--------------------------------+
+| Flash Attention          | 3.0.0.post1                    |
+--------------------------+--------------------------------+
+| hipBLASLt                | 0.15.0-8c6919d                 |
+--------------------------+--------------------------------+
+| Triton                   | 3.3.0                          |
+--------------------------+--------------------------------+
+
+.. _amd-pytorch-training-model-support-v256:
+
+Supported models
+================
+
+The following models are pre-optimized for performance on the AMD Instinct MI325X and MI300X accelerators.
+
+.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/pytorch-training-v25.6-benchmark-models.yaml
+
+   {% set unified_docker = data.unified_docker.latest %}
+   {% set model_groups = data.model_groups %}
+
+   .. raw:: html
+
+      <div id="vllm-benchmark-ud-params-picker" class="container-fluid">
+        <div class="row">
+          <div class="col-2 me-2 model-param-head">Workload</div>
+          <div class="row col-10">
+   {% for model_group in model_groups %}
+            <div class="col-6 model-param" data-param-k="model-group" data-param-v="{{ model_group.tag }}" tabindex="0">{{ model_group.group }}</div>
+   {% endfor %}
+          </div>
+        </div>
+
+        <div class="row mt-1">
+          <div class="col-2 me-2 model-param-head">Model</div>
+          <div class="row col-10">
+   {% for model_group in model_groups %}
+      {% set models = model_group.models %}
+      {% for model in models %}
+         {% if models|length % 3 == 0 %}
+            <div class="col-4 model-param" data-param-k="model" data-param-v="{{ model.mad_tag }}" data-param-group="{{ model_group.tag }}" tabindex="0">{{ model.model }}</div>
+         {% else %}
+            <div class="col-6 model-param" data-param-k="model" data-param-v="{{ model.mad_tag }}" data-param-group="{{ model_group.tag }}" tabindex="0">{{ model.model }}</div>
+         {% endif %}
+      {% endfor %}
+   {% endfor %}
+          </div>
+        </div>
+      </div>
+
+   .. note::
+
+      Some models require an external license agreement through a third party (for example, Meta).
+
+   .. _amd-pytorch-training-performance-measurements-v256:
+
+   Performance measurements
+   ========================
+
+   To evaluate performance, the
+   `Performance results with AMD ROCm software <https://www.amd.com/en/developer/resources/rocm-hub/dev-ai/performance-results.html#tabs-a8deaeb413-item-21cea50186-tab>`_
+   page provides reference throughput and latency measurements for training
+   popular AI models.
+
+   .. note::
+
+      The performance data presented in
+      `Performance results with AMD ROCm software <https://www.amd.com/en/developer/resources/rocm-hub/dev-ai/performance-results.html#tabs-a8deaeb413-item-21cea50186-tab>`_
+      should not be interpreted as the peak performance achievable by AMD
+      Instinct MI325X and MI300X accelerators or ROCm software.
+
+   System validation
+   =================
+
+   Before running AI workloads, it's important to validate that your AMD hardware is configured
+   correctly and performing optimally.
+
+   If you have already validated your system settings, including aspects like NUMA auto-balancing, you
+   can skip this step. Otherwise, complete the procedures in the :ref:`System validation and
+   optimization <rocm-for-ai-system-optimization>` guide to properly configure your system settings
+   before starting training.
+
+   To test for optimal performance, consult the recommended :ref:`System health benchmarks
+   <rocm-for-ai-system-health-bench>`. This suite of tests will help you verify and fine-tune your
+   system's configuration.
+
+   This Docker image is optimized for specific model configurations outlined
+   below. Performance can vary for other training workloads, as AMD
+   doesn’t validate configurations and run conditions outside those described.
+
+   Benchmarking
+   ============
+
+   Once the setup is complete, choose between two options to start benchmarking:
+
+   .. tab-set::
+
+      .. tab-item:: MAD-integrated benchmarking
+
+         Clone the ROCm Model Automation and Dashboarding (`<https://github.com/ROCm/MAD>`__) repository to a local
+         directory and install the required packages on the host machine.
+
+         .. code-block:: shell
+
+            git clone https://github.com/ROCm/MAD
+            cd MAD
+            pip install -r requirements.txt
+
+   {% for model_group in model_groups %}
+      {% for model in model_group.models %}
+
+         .. container:: model-doc {{ model.mad_tag }}
+
+            For example, use this command to run the performance benchmark test on the {{ model.model }} model
+            using one GPU with the {{ model.precision }} data type on the host machine.
+
+            .. code-block:: shell
+
+               export MAD_SECRETS_HFTOKEN="your personal Hugging Face token to access gated models"
+               madengine run \
+                   --tags {{ model.mad_tag }} \
+                   --keep-model-dir \
+                   --live-output \
+                   --timeout 28800
+
+            MAD launches a Docker container with the name
+            ``container_ci-{{ model.mad_tag }}``, for example. The latency and throughput reports of the
+            model are collected in the following path: ``~/MAD/perf.csv``.
+
+      {% endfor %}
+   {% endfor %}
+
+      .. tab-item:: Standalone benchmarking
+
+         .. rubric:: Download the Docker image and required packages
+
+         Use the following command to pull the Docker image from Docker Hub.
+
+         .. code-block:: shell
+
+            docker pull {{ unified_docker.pull_tag }}
+
+         Run the Docker container.
+
+         .. code-block:: shell
+
+            docker run -it --device /dev/dri --device /dev/kfd --network host --ipc host --group-add video --cap-add SYS_PTRACE --security-opt seccomp=unconfined --privileged -v $HOME:$HOME -v  $HOME/.ssh:/root/.ssh --shm-size 64G --name training_env {{ unified_docker.pull_tag }}
+
+         Use these commands if you exit the ``training_env`` container and need to return to it.
+
+         .. code-block:: shell
+
+            docker start training_env
+            docker exec -it training_env bash
+
+         In the Docker container, clone the `<https://github.com/ROCm/MAD>`__
+         repository and navigate to the benchmark scripts directory
+         ``/workspace/MAD/scripts/pytorch_train``.
+
+         .. code-block:: shell
+
+            git clone https://github.com/ROCm/MAD
+            cd MAD/scripts/pytorch_train
+
+         .. rubric:: Prepare training datasets and dependencies
+
+         The following benchmarking examples require downloading models and datasets
+         from Hugging Face. To ensure successful access to gated repos, set your
+         ``HF_TOKEN``.
+
+         .. code-block:: shell
+
+            export HF_TOKEN=$your_personal_hugging_face_access_token
+
+         Run the setup script to install libraries and datasets needed for benchmarking.
+
+         .. code-block:: shell
+
+            ./pytorch_benchmark_setup.sh
+
+         .. container:: model-doc pyt_train_llama-3.1-8b
+
+            ``pytorch_benchmark_setup.sh`` installs the following libraries for Llama 3.1 8B:
+
+            .. list-table::
+               :header-rows: 1
+
+               * - Library
+                 - Reference
+
+               * - ``accelerate``
+                 - `Hugging Face Accelerate <https://huggingface.co/docs/accelerate/en/index>`_
+
+               * - ``datasets``
+                 - `Hugging Face Datasets <https://huggingface.co/docs/datasets/v3.2.0/en/index>`_ 3.2.0
+
+         .. container:: model-doc pyt_train_llama-3.1-70b
+
+            ``pytorch_benchmark_setup.sh`` installs the following libraries for Llama 3.1 70B:
+
+            .. list-table::
+               :header-rows: 1
+
+               * - Library
+                 - Reference
+
+               * - ``datasets``
+                 - `Hugging Face Datasets <https://huggingface.co/docs/datasets/v3.2.0/en/index>`_ 3.2.0
+
+               * - ``torchdata``
+                 - `TorchData <https://pytorch.org/data/beta/index.html>`_
+
+               * - ``tomli``
+                 - `Tomli <https://pypi.org/project/tomli/>`_
+
+               * - ``tiktoken``
+                 - `tiktoken <https://github.com/openai/tiktoken>`_
+
+               * - ``blobfile``
+                 - `blobfile <https://pypi.org/project/blobfile/>`_
+
+               * - ``tabulate``
+                 - `tabulate <https://pypi.org/project/tabulate/>`_
+
+               * - ``wandb``
+                 - `Weights & Biases <https://github.com/wandb/wandb>`_
+
+               * - ``sentencepiece``
+                 - `SentencePiece <https://github.com/google/sentencepiece>`_ 0.2.0
+
+               * - ``tensorboard``
+                 - `TensorBoard <https://www.tensorflow.org/tensorboard>`_ 2.18.0
+
+         .. container:: model-doc pyt_train_flux
+
+            ``pytorch_benchmark_setup.sh`` installs the following libraries for FLUX:
+
+            .. list-table::
+               :header-rows: 1
+
+               * - Library
+                 - Reference
+
+               * - ``accelerate``
+                 - `Hugging Face Accelerate <https://huggingface.co/docs/accelerate/en/index>`_
+
+               * - ``datasets``
+                 - `Hugging Face Datasets <https://huggingface.co/docs/datasets/v3.2.0/en/index>`_ 3.2.0
+
+               * - ``sentencepiece``
+                 - `SentencePiece <https://github.com/google/sentencepiece>`_ 0.2.0
+
+               * - ``tensorboard``
+                 - `TensorBoard <https://www.tensorflow.org/tensorboard>`_ 2.18.0
+
+               * - ``csvkit``
+                 - `csvkit <https://csvkit.readthedocs.io/en/latest/>`_ 2.0.1
+
+               * - ``deepspeed``
+                 - `DeepSpeed <https://github.com/deepspeedai/DeepSpeed>`_ 0.16.2
+
+               * - ``diffusers``
+                 - `Hugging Face Diffusers <https://huggingface.co/docs/diffusers/en/index>`_ 0.31.0
+
+               * - ``GitPython``
+                 - `GitPython <https://github.com/gitpython-developers/GitPython>`_ 3.1.44
+
+               * - ``opencv-python-headless``
+                 - `opencv-python-headless <https://pypi.org/project/opencv-python-headless/>`_ 4.10.0.84
+
+               * - ``peft``
+                 - `PEFT <https://huggingface.co/docs/peft/en/index>`_ 0.14.0
+
+               * - ``protobuf``
+                 - `Protocol Buffers <https://github.com/protocolbuffers/protobuf>`_ 5.29.2
+
+               * - ``pytest``
+                 - `PyTest <https://docs.pytest.org/en/stable/>`_ 8.3.4
+
+               * - ``python-dotenv``
+                 - `python-dotenv <https://pypi.org/project/python-dotenv/>`_ 1.0.1
+
+               * - ``seaborn``
+                 - `Seaborn <https://seaborn.pydata.org/>`_ 0.13.2
+
+               * - ``transformers``
+                 - `Transformers <https://huggingface.co/docs/transformers/en/index>`_ 4.47.0
+
+         ``pytorch_benchmark_setup.sh`` downloads the following datasets from Hugging Face:
+
+         * `bghira/pseudo-camera-10k <https://huggingface.co/datasets/bghira/pseudo-camera-10k>`_
+
+   {% for model_group in model_groups %}
+      {% for model in model_group.models %}
+         {% if model_group.tag == "pre-training" and model.mad_tag in ["pyt_train_llama-3.1-8b", "pyt_train_llama-3.1-70b", "pyt_train_flux"] %}
+
+         .. container:: model-doc {{ model.mad_tag }}
+
+            .. rubric:: Pretraining
+
+            To start the pre-training benchmark, use the following command with the
+            appropriate options. See the following list of options and their descriptions.
+
+            .. code-block:: shell
+
+               ./pytorch_benchmark_report.sh -t pretrain -m {{ model.model_repo }} -p $datatype -s $sequence_length
+
+            .. list-table::
+               :header-rows: 1
+
+               * - Name
+                 - Options
+                 - Description
+
+            {% if model.mad_tag == "pyt_train_llama-3.1-8b" %}
+               * - ``$datatype``
+                 - ``BF16`` or ``FP8``
+                 - Only Llama 3.1 8B supports FP8 precision.
+            {% else %}
+               * - ``$datatype``
+                 - ``BF16``
+                 - Only Llama 3.1 8B supports FP8 precision.
+            {% endif %}
+
+               * - ``$sequence_length``
+                 - Sequence length for the language model.
+                 - Between 2048 and 8192. 8192 by default.
+
+            {% if model.mad_tag == "pyt_train_flux" %}
+            .. container:: model-doc {{ model.mad_tag }}
+
+               .. note::
+
+                  Occasionally, downloading the Flux dataset might fail. In the event of this
+                  error, manually download it from Hugging Face at
+                  `black-forest-labs/FLUX.1-dev <https://huggingface.co/black-forest-labs/FLUX.1-dev>`_
+                  and save it to `/workspace/FluxBenchmark`. This ensures that the test script can access
+                  the required dataset.
+            {% endif %}
+         {% endif %}
+
+         {% if model_group.tag == "fine-tuning" %}
+         .. container:: model-doc {{ model.mad_tag }}
+
+            .. rubric:: Fine-tuning
+
+            To start the fine-tuning benchmark, use the following command with the
+            appropriate options. See the following list of options and their descriptions.
+
+            .. code-block:: shell
+
+               ./pytorch_benchmark_report.sh -t $training_mode -m {{ model.model_repo }} -p BF16 -s $sequence_length
+
+            .. list-table::
+               :header-rows: 1
+
+               * - Name
+                 - Options
+                 - Description
+
+               * - ``$training_mode``
+                 - ``finetune_fw``
+                 - Full weight fine-tuning (BF16 supported)
+
+               * -
+                 - ``finetune_lora``
+                 - LoRA fine-tuning (BF16 supported)
+
+               * -
+                 - ``finetune_qlora``
+                 - QLoRA fine-tuning (BF16 supported)
+
+               * -
+                 - ``HF_finetune_lora``
+                 - LoRA fine-tuning with Hugging Face PEFT
+
+               * - ``$datatype``
+                 - ``BF16``
+                 - All models support BF16.
+
+               * - ``$sequence_length``
+                 - Between 2048 and 16384.
+                 - Sequence length for the language model.
+
+            .. note::
+
+               {{ model.model }} currently supports the following fine-tuning methods:
+
+            {% for method in model.training_modes %}
+               * ``{{ method }}``
+            {% endfor %}
+            {% if model.training_modes|length < 4 %}
+
+               The upstream `torchtune <https://github.com/pytorch/torchtune>`_ repository
+               does not currently provide YAML configuration files for other combinations of
+               model to fine-tuning method
+               However, you can still configure your own YAML files to enable support for
+               fine-tuning methods not listed here by following existing patterns in the
+               ``/workspace/torchtune/recipes/configs`` directory.
+            {% endif %}
+         {% endif %}
+      {% endfor %}
+   {% endfor %}
+
+               .. rubric:: Benchmarking examples
+
+               For examples of benchmarking commands, see `<https://github.com/ROCm/MAD/tree/develop/benchmark/pytorch_train#benchmarking-examples>`__.
+
+Further reading
+===============
+
+- To learn more about MAD and the ``madengine`` CLI, see the `MAD usage guide <https://github.com/ROCm/MAD?tab=readme-ov-file#usage-guide>`__.
+
+- To learn more about system settings and management practices to configure your system for
+  AMD Instinct MI300X series accelerators, see `AMD Instinct MI300X system optimization <https://instinct.docs.amd.com/projects/amdgpu-docs/en/latest/system-optimization/mi300x.html>`_.
+
+- For a list of other ready-made Docker images for AI with ROCm, see
+  `AMD Infinity Hub <https://www.amd.com/en/developer/resources/infinity-hub.html#f-amd_hub_category=AI%20%26%20ML%20Models>`_.
+
+Previous versions
+=================
+
+See :doc:`pytorch-training-history` to find documentation for previous releases
+of the ``ROCm/pytorch-training`` Docker image.
--- a/docs/how-to/rocm-for-ai/training/benchmark-docker/pytorch-training.rst
+++ b/docs/how-to/rocm-for-ai/training/benchmark-docker/pytorch-training.rst
@@ -9,28 +9,25 @@ Training a model with PyTorch for ROCm
 PyTorch is an open-source machine learning framework that is widely used for
 model training with GPU-optimized components for transformer-based models.

-The `PyTorch for ROCm training Docker <https://hub.docker.com/r/rocm/pytorch-training/tags>`_
-(``rocm/pytorch-training:v25.6``) image provides a prebuilt optimized environment for fine-tuning and pretraining a
-model on AMD Instinct MI325X and MI300X accelerators. It includes the following software components to accelerate
-training workloads:
+.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/pytorch-training-benchmark-models.yaml

-+--------------------------+--------------------------------+
-| Software component       | Version                        |
-+==========================+================================+
-| ROCm                     | 6.3.4                          |
-+--------------------------+--------------------------------+
-| PyTorch                  | 2.8.0a0+git7d205b2             |
-+--------------------------+--------------------------------+
-| Python                   | 3.10.17                        |
-+--------------------------+--------------------------------+
-| Transformer Engine       | 1.14.0+2f85f5f2                |
-+--------------------------+--------------------------------+
-| Flash Attention          | 3.0.0.post1                    |
-+--------------------------+--------------------------------+
-| hipBLASLt                | 0.15.0-8c6919d                 |
-+--------------------------+--------------------------------+
-| Triton                   | 3.3.0                          |
-+--------------------------+--------------------------------+
+   {% set dockers = data.dockers %}
+   {% set docker = dockers[0] %}
+   The `PyTorch for ROCm training Docker <{{ docker.docker_hub_url }}>`__
+   (``{{ docker.pull_tag }}``) image provides a prebuilt optimized environment for fine-tuning and pretraining a
+   model on AMD Instinct MI325X and MI300X accelerators. It includes the following software components to accelerate
+   training workloads:
+
+   .. list-table::
+      :header-rows: 1
+
+      * - Software component
+        - Version
+
+      {% for component_name, component_version in docker.components.items() %}
+      * - {{ component_name }}
+        - {{ component_version }}
+      {% endfor %}

 .. _amd-pytorch-training-model-support:

@@ -38,26 +35,27 @@ Supported models
 ================

 The following models are pre-optimized for performance on the AMD Instinct MI325X and MI300X accelerators.
+Some instructions, commands, and training recommendations in this documentation might
+vary by model -- select one to get started.

 .. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/pytorch-training-benchmark-models.yaml

-   {% set unified_docker = data.unified_docker.latest %}
+   {% set unified_docker = data.dockers[0] %}
   {% set model_groups = data.model_groups %}
-
   .. raw:: html

      <div id="vllm-benchmark-ud-params-picker" class="container-fluid">
        <div class="row">
-          <div class="col-2 me-2 model-param-head">Workload</div>
+          <div class="col-2 me-2 model-param-head">Model group</div>
          <div class="row col-10">
   {% for model_group in model_groups %}
-            <div class="col-6 model-param" data-param-k="model-group" data-param-v="{{ model_group.tag }}" tabindex="0">{{ model_group.group }}</div>
+            <div class="col-3 model-param" data-param-k="model-group" data-param-v="{{ model_group.tag }}" tabindex="0">{{ model_group.group }}</div>
   {% endfor %}
          </div>
        </div>

        <div class="row mt-1">
-          <div class="col-2 me-2 model-param-head">Model</div>
+          <div class="col-2 me-2 model-param-head">Model variant</div>
          <div class="row col-10">
   {% for model_group in model_groups %}
      {% set models = model_group.models %}
@@ -73,84 +71,116 @@ The following models are pre-optimized for performance on the AMD Instinct MI325
        </div>
      </div>

-   .. note::

-      Some models require an external license agreement through a third party (for example, Meta).
+   .. _amd-pytorch-training-supported-training-modes:

-   .. _amd-pytorch-training-performance-measurements:
+   The following table lists supported training modes per model.

-   Performance measurements
-   ========================
+   .. dropdown:: Supported training modes

-   To evaluate performance, the
+      .. list-table::
+         :header-rows: 1
+
+         * - Model
+           - Supported training modes
+
+      {% for model_group in model_groups %}
+         {% set models = model_group.models %}
+         {% for model in models %}
+         * - {{ model.model }}
+           - ``{{ model.training_modes | join('``, ``') }}``
+
+         {% endfor %}
+      {% endfor %}
+
+      .. note::
+
+         Some model and fine-tuning combinations are not listed. This is
+         because the `upstream torchtune repository <https://github.com/pytorch/torchtune>`__
+         doesn't provide default YAML configurations for them.
+         For advanced usage, you can create a custom configuration to enable
+         unlisted fine-tuning methods by using an existing file in the
+         ``/workspace/torchtune/recipes/configs`` directory as a template.
+
+.. _amd-pytorch-training-performance-measurements:
+
+Performance measurements
+========================
+
+To evaluate performance, the
+`Performance results with AMD ROCm software <https://www.amd.com/en/developer/resources/rocm-hub/dev-ai/performance-results.html#tabs-a8deaeb413-item-21cea50186-tab>`_
+page provides reference throughput and latency measurements for training
+popular AI models.
+
+.. note::
+
+   The performance data presented in
   `Performance results with AMD ROCm software <https://www.amd.com/en/developer/resources/rocm-hub/dev-ai/performance-results.html#tabs-a8deaeb413-item-21cea50186-tab>`_
-   page provides reference throughput and latency measurements for training
-   popular AI models.
+   should not be interpreted as the peak performance achievable by AMD
+   Instinct MI325X and MI300X accelerators or ROCm software.

-   .. note::
+System validation
+=================

-      The performance data presented in
-      `Performance results with AMD ROCm software <https://www.amd.com/en/developer/resources/rocm-hub/dev-ai/performance-results.html#tabs-a8deaeb413-item-21cea50186-tab>`_
-      should not be interpreted as the peak performance achievable by AMD
-      Instinct MI325X and MI300X accelerators or ROCm software.
+Before running AI workloads, it's important to validate that your AMD hardware is configured
+correctly and performing optimally.

-   System validation
-   =================
+If you have already validated your system settings, including aspects like NUMA auto-balancing, you
+can skip this step. Otherwise, complete the procedures in the :ref:`System validation and
+optimization <rocm-for-ai-system-optimization>` guide to properly configure your system settings
+before starting training.

-   Before running AI workloads, it's important to validate that your AMD hardware is configured
-   correctly and performing optimally.
+To test for optimal performance, consult the recommended :ref:`System health benchmarks
+<rocm-for-ai-system-health-bench>`. This suite of tests will help you verify and fine-tune your
+system's configuration.

-   If you have already validated your system settings, including aspects like NUMA auto-balancing, you
-   can skip this step. Otherwise, complete the procedures in the :ref:`System validation and
-   optimization <rocm-for-ai-system-optimization>` guide to properly configure your system settings
-   before starting training.
+This Docker image is optimized for specific model configurations outlined
+below. Performance can vary for other training workloads, as AMD
+doesn’t test configurations and run conditions outside those described.

-   To test for optimal performance, consult the recommended :ref:`System health benchmarks
-   <rocm-for-ai-system-health-bench>`. This suite of tests will help you verify and fine-tune your
-   system's configuration.
+Run training
+============

-   This Docker image is optimized for specific model configurations outlined
-   below. Performance can vary for other training workloads, as AMD
-   doesn’t validate configurations and run conditions outside those described.
+.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/pytorch-training-benchmark-models.yaml

-   Benchmarking
-   ============
+   {% set unified_docker = data.dockers[0] %}
+   {% set model_groups = data.model_groups %}

-   Once the setup is complete, choose between two options to start benchmarking:
+   Once the setup is complete, choose between two options to start benchmarking training:

   .. tab-set::

      .. tab-item:: MAD-integrated benchmarking

-         Clone the ROCm Model Automation and Dashboarding (`<https://github.com/ROCm/MAD>`__) repository to a local
-         directory and install the required packages on the host machine.
+         1. Clone the ROCm Model Automation and Dashboarding (`<https://github.com/ROCm/MAD>`__) repository to a local
+            directory and install the required packages on the host machine.

-         .. code-block:: shell
+            .. code-block:: shell

-            git clone https://github.com/ROCm/MAD
-            cd MAD
-            pip install -r requirements.txt
+               git clone https://github.com/ROCm/MAD
+               cd MAD
+               pip install -r requirements.txt

   {% for model_group in model_groups %}
      {% for model in model_group.models %}

         .. container:: model-doc {{ model.mad_tag }}

-            For example, use this command to run the performance benchmark test on the {{ model.model }} model
-            using one GPU with the {{ model.precision }} data type on the host machine.
+            2. For example, use this command to run the performance benchmark test on the {{ model.model }} model
+               using one node with the {{ model.precision }} data type on the host machine.

-            .. code-block:: shell
+               .. code-block:: shell

-               export MAD_SECRETS_HFTOKEN="your personal Hugging Face token to access gated models"
-               madengine run \
-                   --tags {{ model.mad_tag }} \
-                   --keep-model-dir \
-                   --live-output \
-                   --timeout 28800
+                  export MAD_SECRETS_HFTOKEN="your personal Hugging Face token to access gated models"
+                  madengine run \
+                      --tags {{ model.mad_tag }} \
+                      --keep-model-dir \
+                      --live-output \
+                      --timeout 28800

-            MAD launches a Docker container with the name
-            ``container_ci-{{ model.mad_tag }}``, for example. The latency and throughput reports of the
-            model are collected in the following path: ``~/MAD/perf.csv``.
+               MAD launches a Docker container with the name
+               ``container_ci-{{ model.mad_tag }}``. The latency and throughput reports of the
+               model are collected in ``~/MAD/perf.csv``.

      {% endfor %}
   {% endfor %}
@@ -159,222 +189,213 @@ The following models are pre-optimized for performance on the AMD Instinct MI325

         .. rubric:: Download the Docker image and required packages

-         Use the following command to pull the Docker image from Docker Hub.
+         1. Use the following command to pull the Docker image from Docker Hub.

-         .. code-block:: shell
+            .. code-block:: shell

-            docker pull {{ unified_docker.pull_tag }}
+               docker pull {{ unified_docker.pull_tag }}

-         Run the Docker container.
+         2. Run the Docker container.

-         .. code-block:: shell
+            .. code-block:: shell

-            docker run -it --device /dev/dri --device /dev/kfd --network host --ipc host --group-add video --cap-add SYS_PTRACE --security-opt seccomp=unconfined --privileged -v $HOME:$HOME -v  $HOME/.ssh:/root/.ssh --shm-size 64G --name training_env {{ unified_docker.pull_tag }}
+               docker run -it \
+                   --device /dev/dri \
+                   --device /dev/kfd \
+                   --network host \
+                   --ipc host \
+                   --group-add video \
+                   --cap-add SYS_PTRACE \
+                   --security-opt seccomp=unconfined \
+                   --privileged \
+                   -v $HOME:$HOME \
+                   -v $HOME/.ssh:/root/.ssh \
+                   --shm-size 64G \
+                   --name training_env \
+                   {{ unified_docker.pull_tag }}

-         Use these commands if you exit the ``training_env`` container and need to return to it.
+            Use these commands if you exit the ``training_env`` container and need to return to it.

-         .. code-block:: shell
+            .. code-block:: shell

-            docker start training_env
-            docker exec -it training_env bash
+               docker start training_env
+               docker exec -it training_env bash

-         In the Docker container, clone the `<https://github.com/ROCm/MAD>`__
-         repository and navigate to the benchmark scripts directory
-         ``/workspace/MAD/scripts/pytorch_train``.
+         3. In the Docker container, clone the `<https://github.com/ROCm/MAD>`__
+            repository and navigate to the benchmark scripts directory
+            ``/workspace/MAD/scripts/pytorch_train``.

-         .. code-block:: shell
+            .. code-block:: shell

-            git clone https://github.com/ROCm/MAD
-            cd MAD/scripts/pytorch_train
+               git clone https://github.com/ROCm/MAD
+               cd MAD/scripts/pytorch_train

         .. rubric:: Prepare training datasets and dependencies

-         The following benchmarking examples require downloading models and datasets
-         from Hugging Face. To ensure successful access to gated repos, set your
-         ``HF_TOKEN``.
+         1. The following benchmarking examples require downloading models and datasets
+            from Hugging Face. To ensure successful access to gated repos, set your
+            ``HF_TOKEN``.

-         .. code-block:: shell
+            .. code-block:: shell

-            export HF_TOKEN=$your_personal_hugging_face_access_token
+               export HF_TOKEN=$your_personal_hugging_face_access_token

-         Run the setup script to install libraries and datasets needed for benchmarking.
+         2. Run the setup script to install libraries and datasets needed for benchmarking.

-         .. code-block:: shell
+            .. code-block:: shell

-            ./pytorch_benchmark_setup.sh
+               ./pytorch_benchmark_setup.sh

-         .. container:: model-doc pyt_train_llama-3.1-8b
+            .. container:: model-doc pyt_train_llama-3.1-8b

-            ``pytorch_benchmark_setup.sh`` installs the following libraries for Llama 3.1 8B:
+               ``pytorch_benchmark_setup.sh`` installs the following libraries for Llama 3.1 8B:

-            .. list-table::
-               :header-rows: 1
+               .. list-table::
+                  :header-rows: 1

-               * - Library
-                 - Reference
+                  * - Library
+                    - Reference

-               * - ``accelerate``
-                 - `Hugging Face Accelerate <https://huggingface.co/docs/accelerate/en/index>`_
+                  * - ``accelerate``
+                    - `Hugging Face Accelerate <https://huggingface.co/docs/accelerate/en/index>`_

-               * - ``datasets``
-                 - `Hugging Face Datasets <https://huggingface.co/docs/datasets/v3.2.0/en/index>`_ 3.2.0
+                  * - ``datasets``
+                    - `Hugging Face Datasets <https://huggingface.co/docs/datasets/v3.2.0/en/index>`_ 3.2.0

-         .. container:: model-doc pyt_train_llama-3.1-70b
+            .. container:: model-doc pyt_train_llama-3.1-70b

-            ``pytorch_benchmark_setup.sh`` installs the following libraries for Llama 3.1 70B:
+               ``pytorch_benchmark_setup.sh`` installs the following libraries for Llama 3.1 70B:

-            .. list-table::
-               :header-rows: 1
+               .. list-table::
+                  :header-rows: 1

-               * - Library
-                 - Reference
+                  * - Library
+                    - Reference

-               * - ``datasets``
-                 - `Hugging Face Datasets <https://huggingface.co/docs/datasets/v3.2.0/en/index>`_ 3.2.0
+                  * - ``datasets``
+                    - `Hugging Face Datasets <https://huggingface.co/docs/datasets/v3.2.0/en/index>`_ 3.2.0

-               * - ``torchdata``
-                 - `TorchData <https://pytorch.org/data/beta/index.html>`_
+                  * - ``torchdata``
+                    - `TorchData <https://pytorch.org/data/beta/index.html>`_

-               * - ``tomli``
-                 - `Tomli <https://pypi.org/project/tomli/>`_
+                  * - ``tomli``
+                    - `Tomli <https://pypi.org/project/tomli/>`_

-               * - ``tiktoken``
-                 - `tiktoken <https://github.com/openai/tiktoken>`_
+                  * - ``tiktoken``
+                    - `tiktoken <https://github.com/openai/tiktoken>`_

-               * - ``blobfile``
-                 - `blobfile <https://pypi.org/project/blobfile/>`_
+                  * - ``blobfile``
+                    - `blobfile <https://pypi.org/project/blobfile/>`_

-               * - ``tabulate``
-                 - `tabulate <https://pypi.org/project/tabulate/>`_
+                  * - ``tabulate``
+                    - `tabulate <https://pypi.org/project/tabulate/>`_

-               * - ``wandb``
-                 - `Weights & Biases <https://github.com/wandb/wandb>`_
+                  * - ``wandb``
+                    - `Weights & Biases <https://github.com/wandb/wandb>`_

-               * - ``sentencepiece``
-                 - `SentencePiece <https://github.com/google/sentencepiece>`_ 0.2.0
+                  * - ``sentencepiece``
+                    - `SentencePiece <https://github.com/google/sentencepiece>`_ 0.2.0

-               * - ``tensorboard``
-                 - `TensorBoard <https://www.tensorflow.org/tensorboard>`_ 2.18.0
+                  * - ``tensorboard``
+                    - `TensorBoard <https://www.tensorflow.org/tensorboard>`_ 2.18.0

-         .. container:: model-doc pyt_train_flux
+            .. container:: model-doc pyt_train_flux

-            ``pytorch_benchmark_setup.sh`` installs the following libraries for FLUX:
+               ``pytorch_benchmark_setup.sh`` installs the following libraries for FLUX:

-            .. list-table::
-               :header-rows: 1
+               .. list-table::
+                  :header-rows: 1

-               * - Library
-                 - Reference
+                  * - Library
+                    - Reference

-               * - ``accelerate``
-                 - `Hugging Face Accelerate <https://huggingface.co/docs/accelerate/en/index>`_
+                  * - ``accelerate``
+                    - `Hugging Face Accelerate <https://huggingface.co/docs/accelerate/en/index>`_

-               * - ``datasets``
-                 - `Hugging Face Datasets <https://huggingface.co/docs/datasets/v3.2.0/en/index>`_ 3.2.0
+                  * - ``datasets``
+                    - `Hugging Face Datasets <https://huggingface.co/docs/datasets/v3.2.0/en/index>`_ 3.2.0

-               * - ``sentencepiece``
-                 - `SentencePiece <https://github.com/google/sentencepiece>`_ 0.2.0
+                  * - ``sentencepiece``
+                    - `SentencePiece <https://github.com/google/sentencepiece>`_ 0.2.0

-               * - ``tensorboard``
-                 - `TensorBoard <https://www.tensorflow.org/tensorboard>`_ 2.18.0
+                  * - ``tensorboard``
+                    - `TensorBoard <https://www.tensorflow.org/tensorboard>`_ 2.18.0

-               * - ``csvkit``
-                 - `csvkit <https://csvkit.readthedocs.io/en/latest/>`_ 2.0.1
+                  * - ``csvkit``
+                    - `csvkit <https://csvkit.readthedocs.io/en/latest/>`_ 2.0.1

-               * - ``deepspeed``
-                 - `DeepSpeed <https://github.com/deepspeedai/DeepSpeed>`_ 0.16.2
+                  * - ``deepspeed``
+                    - `DeepSpeed <https://github.com/deepspeedai/DeepSpeed>`_ 0.16.2

-               * - ``diffusers``
-                 - `Hugging Face Diffusers <https://huggingface.co/docs/diffusers/en/index>`_ 0.31.0
+                  * - ``diffusers``
+                    - `Hugging Face Diffusers <https://huggingface.co/docs/diffusers/en/index>`_ 0.31.0

-               * - ``GitPython``
-                 - `GitPython <https://github.com/gitpython-developers/GitPython>`_ 3.1.44
+                  * - ``GitPython``
+                    - `GitPython <https://github.com/gitpython-developers/GitPython>`_ 3.1.44

-               * - ``opencv-python-headless``
-                 - `opencv-python-headless <https://pypi.org/project/opencv-python-headless/>`_ 4.10.0.84
+                  * - ``opencv-python-headless``
+                    - `opencv-python-headless <https://pypi.org/project/opencv-python-headless/>`_ 4.10.0.84

-               * - ``peft``
-                 - `PEFT <https://huggingface.co/docs/peft/en/index>`_ 0.14.0
+                  * - ``peft``
+                    - `PEFT <https://huggingface.co/docs/peft/en/index>`_ 0.14.0

-               * - ``protobuf``
-                 - `Protocol Buffers <https://github.com/protocolbuffers/protobuf>`_ 5.29.2
+                  * - ``protobuf``
+                    - `Protocol Buffers <https://github.com/protocolbuffers/protobuf>`_ 5.29.2

-               * - ``pytest``
-                 - `PyTest <https://docs.pytest.org/en/stable/>`_ 8.3.4
+                  * - ``pytest``
+                    - `PyTest <https://docs.pytest.org/en/stable/>`_ 8.3.4

-               * - ``python-dotenv``
-                 - `python-dotenv <https://pypi.org/project/python-dotenv/>`_ 1.0.1
+                  * - ``python-dotenv``
+                    - `python-dotenv <https://pypi.org/project/python-dotenv/>`_ 1.0.1

-               * - ``seaborn``
-                 - `Seaborn <https://seaborn.pydata.org/>`_ 0.13.2
+                  * - ``seaborn``
+                    - `Seaborn <https://seaborn.pydata.org/>`_ 0.13.2

-               * - ``transformers``
-                 - `Transformers <https://huggingface.co/docs/transformers/en/index>`_ 4.47.0
+                  * - ``transformers``
+                    - `Transformers <https://huggingface.co/docs/transformers/en/index>`_ 4.47.0

-         ``pytorch_benchmark_setup.sh`` downloads the following datasets from Hugging Face:
+            ``pytorch_benchmark_setup.sh`` downloads the following datasets from Hugging Face:

-         * `bghira/pseudo-camera-10k <https://huggingface.co/datasets/bghira/pseudo-camera-10k>`_
+            * `bghira/pseudo-camera-10k <https://huggingface.co/datasets/bghira/pseudo-camera-10k>`_

   {% for model_group in model_groups %}
      {% for model in model_group.models %}
-         {% if model_group.tag == "pre-training" and model.mad_tag in ["pyt_train_llama-3.1-8b", "pyt_train_llama-3.1-70b", "pyt_train_flux"] %}
+         {% set training_modes = model.training_modes %}
+         {% set training_mode_descs = {
+            "pretrain": "Benchmark pre-training.",
+            "HF_pretrain": "Llama 3.1 8B pre-training with FP8 precision."
+         } %}
+         {% set available_modes = training_modes | select("in", ["pretrain", "HF_pretrain"]) | list %}
+         {% if available_modes %}

         .. container:: model-doc {{ model.mad_tag }}

-            .. rubric:: Pretraining
+            .. rubric:: Pre-training

            To start the pre-training benchmark, use the following command with the
            appropriate options. See the following list of options and their descriptions.

            .. code-block:: shell

-               ./pytorch_benchmark_report.sh -t pretrain -m {{ model.model_repo }} -p $datatype -s $sequence_length
-
-            .. list-table::
-               :header-rows: 1
-
-               * - Name
-                 - Options
-                 - Description
-
-            {% if model.mad_tag == "pyt_train_llama-3.1-8b" %}
-               * - ``$datatype``
-                 - ``BF16`` or ``FP8``
-                 - Only Llama 3.1 8B supports FP8 precision.
-            {% else %}
-               * - ``$datatype``
-                 - ``BF16``
-                 - Only Llama 3.1 8B supports FP8 precision.
-            {% endif %}
-
-               * - ``$sequence_length``
-                 - Sequence length for the language model.
-                 - Between 2048 and 8192. 8192 by default.
+               ./pytorch_benchmark_report.sh -t {% if available_modes | length == 1 %}{{ available_modes[0] }}{% else %}$training_mode{% endif %} \
+                   -m {{ model.model_repo }} \
+                   -p $datatype \
+                   -s $sequence_length

            {% if model.mad_tag == "pyt_train_flux" %}
            .. container:: model-doc {{ model.mad_tag }}

               .. note::

+                  Currently, FLUX models are not supported out-of-the-box on {{ unified_docker.pull_tag }}.
+                  To use FLUX, refer to the previous version of the ``pytorch-training`` Docker: :doc:`previous-versions/pytorch-training-v25.6`
+
                  Occasionally, downloading the Flux dataset might fail. In the event of this
                  error, manually download it from Hugging Face at
                  `black-forest-labs/FLUX.1-dev <https://huggingface.co/black-forest-labs/FLUX.1-dev>`_
                  and save it to `/workspace/FluxBenchmark`. This ensures that the test script can access
                  the required dataset.
            {% endif %}
-         {% endif %}
-
-         {% if model_group.tag == "fine-tuning" %}
-         .. container:: model-doc {{ model.mad_tag }}
-
-            .. rubric:: Fine-tuning
-
-            To start the fine-tuning benchmark, use the following command with the
-            appropriate options. See the following list of options and their descriptions.
-
-            .. code-block:: shell
-
-               ./pytorch_benchmark_report.sh -t $training_mode -m {{ model.model_repo }} -p BF16 -s $sequence_length

            .. list-table::
               :header-rows: 1
@@ -383,53 +404,143 @@ The following models are pre-optimized for performance on the AMD Instinct MI325
                 - Options
                 - Description

-               * - ``$training_mode``
-                 - ``finetune_fw``
-                 - Full weight fine-tuning (BF16 supported)
-
-               * -
-                 - ``finetune_lora``
-                 - LoRA fine-tuning (BF16 supported)
-
-               * -
-                 - ``finetune_qlora``
-                 - QLoRA fine-tuning (BF16 supported)
-
-               * -
-                 - ``HF_finetune_lora``
-                 - LoRA fine-tuning with Hugging Face PEFT
+               {% for mode in available_modes %}
+               * - {% if loop.first %}``$training_mode``{% endif %}
+                 - ``{{ mode }}``
+                 - {{ training_mode_descs[mode] }}
+               {% endfor %}

               * - ``$datatype``
-                 - ``BF16``
-                 - All models support BF16.
+                 - ``BF16``{% if model.mad_tag == "pyt_train_llama-3.1-8b" %} or ``FP8``{% endif %}
+                 - Only Llama 3.1 8B supports FP8 precision.
+
+               * - ``$sequence_length``
+                 - Sequence length for the language model.
+                 - Between 2048 and 8192. 8192 by default.
+         {% endif %}
+
+         {% set training_mode_descs = {
+            "finetune_fw": "Full weight fine-tuning (BF16 and FP8 supported).",
+            "finetune_lora": "LoRA fine-tuning (BF16 supported).",
+            "finetune_qlora": "QLoRA fine-tuning (BF16 supported).",
+            "HF_finetune_lora": "LoRA fine-tuning with Hugging Face PEFT.",
+         } %}
+         {% set available_modes = training_modes | select("in", ["finetune_fw", "finetune_lora", "finetune_qlora", "HF_finetune_lora"]) | list %}
+         {% if available_modes %}
+         .. container:: model-doc {{ model.mad_tag }}
+
+            .. rubric:: Fine-tuning
+
+            To start the fine-tuning benchmark, use the following command with the
+            appropriate options. See the following list of options and their descriptions.
+            See :ref:`supported training modes <amd-pytorch-training-supported-training-modes>`.
+
+            .. code-block:: shell
+
+               ./pytorch_benchmark_report.sh -t $training_mode \
+                   -m {{ model.model_repo }} \
+                   -p $datatype \
+                   -s $sequence_length
+
+            .. list-table::
+               :header-rows: 1
+
+               * - Name
+                 - Options
+                 - Description
+
+               {% for mode in available_modes %}
+               * - {% if loop.first %}``$training_mode``{% endif %}
+                 - ``{{ mode }}``
+                 - {{ training_mode_descs[mode] }}
+               {% endfor %}
+
+               * - ``$datatype``
+                 - ``BF16``{% if "finetune_fw" in available_modes %} or ``FP8``{% endif %}
+                 - All models support BF16.{% if "finetune_fw" in available_modes %} FP8 is only available for full weight fine-tuning.{% endif %}

               * - ``$sequence_length``
                 - Between 2048 and 16384.
                 - Sequence length for the language model.

+            {% if model.mad_tag in ["pyt_train_llama3.2-vision-11b", "pyt_train_llama-3.2-vision-90b"] %}
            .. note::

-               {{ model.model }} currently supports the following fine-tuning methods:
+               For LoRA and QLoRA support with vision models (Llama 3.2 11B and 90B),
+               use the following torchtune commit for compatibility:

-            {% for method in model.training_modes %}
-               * ``{{ method }}``
-            {% endfor %}
-            {% if model.training_modes|length < 4 %}
+               .. code-block:: shell
+
+                  git checkout 48192e23188b1fc524dd6d127725ceb2348e7f0e
+
+            {% elif model.mad_tag in ["pyt_train_llama-2-7b", "pyt_train_llama-2-13b", "pyt_train_llama-2-70b"] %}
+            .. note::
+
+               You might encounter the following error with Llama 2: ``ValueError: seq_len (16384) of
+               input tensor should be smaller than max_seq_len (4096)``.
+               This error indicates that an input sequence is longer than the model's maximum context window.
+
+               Ensure your tokenized input does not exceed the model's ``max_seq_len`` (4096
+               tokens in this case). You can resolve this by truncating the input or splitting
+               it into smaller chunks before passing it to the model.
+
+               Note on reproducibility: The results in this guide are based on
+               commit ``b4c98ac`` from the upstream
+               `<https://github.com/pytorch/torchtune>`__ repository. For the
+               latest updates, you can use the main branch.

-               The upstream `torchtune <https://github.com/pytorch/torchtune>`_ repository
-               does not currently provide YAML configuration files for other combinations of
-               model to fine-tuning method
-               However, you can still configure your own YAML files to enable support for
-               fine-tuning methods not listed here by following existing patterns in the
-               ``/workspace/torchtune/recipes/configs`` directory.
            {% endif %}
         {% endif %}
      {% endfor %}
   {% endfor %}

-               .. rubric:: Benchmarking examples
+            .. rubric:: Benchmarking examples

-               For examples of benchmarking commands, see `<https://github.com/ROCm/MAD/tree/develop/benchmark/pytorch_train#benchmarking-examples>`__.
+            For examples of benchmarking commands, see `<https://github.com/ROCm/MAD/tree/develop/benchmark/pytorch_train#benchmarking-examples>`__.
+
+Multi-node training
+-------------------
+
+Pre-training
+~~~~~~~~~~~~
+
+Multi-node training with torchtitan is supported. The provided SLURM script is pre-configured for Llama 3 70B.
+
+To launch the training job on a SLURM cluster for Llama 3 70B, run the following commands from the MAD repository.
+
+.. code-block:: shell
+
+   # In the MAD repository
+   cd scripts/pytorch_train
+   sbatch run_slurm_train.sh
+
+Fine-tuning
+~~~~~~~~~~~
+
+Multi-node training with torchtune is supported. The provided SLURM script is pre-configured for Llama 3.3 70B.
+
+To launch the training job on a SLURM cluster for Llama 3.3 70B, run the following commands from the MAD repository.
+
+.. code-block:: shell
+
+   huggingface-cli login # Get access to HF Llama model space
+   huggingface-cli download meta-llama/Llama-3.3-70B-Instruct --local-dir ./models/Llama-3.3-70B-Instruct # Download the Llama 3.3 model locally
+   # In the MAD repository
+   cd scripts/pytorch_train
+   sbatch Torchtune_Multinode.sh
+
+.. note::
+
+   Information regarding benchmark setup:
+
+   * By default, Llama 3.3 70B is fine-tuned using ``alpaca_dataset``.
+   * You can adjust the torchtune `YAML configuration file
+     <https://github.com/pytorch/torchtune/blob/main/recipes/configs/llama3_3/70B_full_multinode.yaml>`__
+     if you're using a different model.
+   * The number of nodes and other parameters can be tuned in the SLURM script ``Torchtune_Multinode.sh``.
+   * Set the ``mounting_paths`` inside the SLURM script.
+
+Once the run is finished, you can find the log files in the ``result_torchtune/`` directory.

 Further reading
 ===============
--- a/docs/sphinx/requirements.in
+++ b/docs/sphinx/requirements.in
@@ -1,4 +1,4 @@
-rocm-docs-core==1.20.1
+rocm-docs-core==1.22.0
 sphinx-reredirects
 sphinx-sitemap
 sphinxcontrib.datatemplates==0.11.0
--- a/docs/sphinx/requirements.txt
+++ b/docs/sphinx/requirements.txt
@@ -23,7 +23,7 @@ beautifulsoup4==4.13.4
    # via pydata-sphinx-theme
 breathe==4.36.0
    # via rocm-docs-core
-certifi==2025.4.26
+certifi==2025.7.14
    # via requests
 cffi==1.17.1
    # via
@@ -35,18 +35,16 @@ click==8.2.1
    # via
    #   jupyter-cache
    #   sphinx-external-toc
-comm==0.2.2
+comm==0.2.3
    # via ipykernel
-cryptography==45.0.3
+cryptography==45.0.5
    # via pyjwt
-debugpy==1.8.14
+debugpy==1.8.15
    # via ipykernel
 decorator==5.2.1
    # via ipython
 defusedxml==0.7.1
    # via sphinxcontrib-datatemplates
-deprecated==1.2.18
-    # via pygithub
 docutils==0.21.2
    # via
    #   myst-parser
@@ -62,7 +60,7 @@ fastjsonschema==2.21.1
    #   rocm-docs-core
 gitdb==4.0.12
    # via gitpython
-gitpython==3.1.44
+gitpython==3.1.45
    # via rocm-docs-core
 greenlet==3.2.3
    # via sqlalchemy
@@ -74,7 +72,7 @@ importlib-metadata==8.7.0
    # via
    #   jupyter-cache
    #   myst-nb
-ipykernel==6.29.5
+ipykernel==6.30.0
    # via myst-nb
 ipython==8.37.0
    # via
@@ -86,7 +84,7 @@ jinja2==3.1.6
    # via
    #   myst-parser
    #   sphinx
-jsonschema==4.24.0
+jsonschema==4.25.0
    # via nbformat
 jsonschema-specifications==2025.4.1
    # via jsonschema
@@ -116,7 +114,7 @@ mdit-py-plugins==0.4.2
    # via myst-parser
 mdurl==0.1.2
    # via markdown-it-py
-myst-nb==1.2.0
+myst-nb==1.3.0
    # via rocm-docs-core
 myst-parser==4.0.1
    # via myst-nb
@@ -134,7 +132,6 @@ nest-asyncio==1.6.0
 packaging==25.0
    # via
    #   ipykernel
-    #   pydata-sphinx-theme
    #   sphinx
 parso==0.8.4
    # via jedi
@@ -152,13 +149,13 @@ pure-eval==0.2.3
    # via stack-data
 pycparser==2.22
    # via cffi
-pydata-sphinx-theme==0.15.4
+pydata-sphinx-theme==0.16.1
    # via
    #   rocm-docs-core
    #   sphinx-book-theme
-pygithub==2.6.1
+pygithub==2.7.0
    # via rocm-docs-core
-pygments==2.19.1
+pygments==2.19.2
    # via
    #   accessible-pygments
    #   ipython
@@ -178,7 +175,7 @@ pyyaml==6.0.2
    #   rocm-docs-core
    #   sphinx-external-toc
    #   sphinxcontrib-datatemplates
-pyzmq==26.4.0
+pyzmq==27.0.0
    # via
    #   ipykernel
    #   jupyter-client
@@ -190,9 +187,9 @@ requests==2.32.4
    # via
    #   pygithub
    #   sphinx
-rocm-docs-core==1.20.1
+rocm-docs-core==1.22.0
    # via -r requirements.in
-rpds-py==0.25.1
+rpds-py==0.26.0
    # via
    #   jsonschema
    #   referencing
@@ -220,7 +217,7 @@ sphinx==8.1.3
    #   sphinx-reredirects
    #   sphinxcontrib-datatemplates
    #   sphinxcontrib-runcmd
-sphinx-book-theme==1.1.4
+sphinx-book-theme==1.1.3
    # via rocm-docs-core
 sphinx-copybutton==0.5.2
    # via rocm-docs-core
@@ -234,7 +231,7 @@ sphinx-notfound-page==1.1.0
    # via rocm-docs-core
 sphinx-reredirects==0.1.6
    # via -r requirements.in
-sphinx-sitemap==2.8.0
+sphinx-sitemap==2.7.2
    # via -r requirements.in
 sphinxcontrib-applehelp==2.0.0
    # via sphinx
@@ -252,7 +249,7 @@ sphinxcontrib-runcmd==0.2.0
    # via sphinxcontrib-datatemplates
 sphinxcontrib-serializinghtml==2.0.0
    # via sphinx
-sqlalchemy==2.0.41
+sqlalchemy==2.0.42
    # via jupyter-cache
 stack-data==0.6.3
    # via ipython
@@ -266,7 +263,6 @@ tornado==6.5.1
    #   jupyter-client
 traitlets==5.14.3
    # via
-    #   comm
    #   ipykernel
    #   ipython
    #   jupyter-client
@@ -274,7 +270,7 @@ traitlets==5.14.3
    #   matplotlib-inline
    #   nbclient
    #   nbformat
-typing-extensions==4.14.0
+typing-extensions==4.14.1
    # via
    #   beautifulsoup4
    #   exceptiongroup
@@ -290,7 +286,5 @@ urllib3==2.5.0
    #   requests
 wcwidth==0.2.13
    # via prompt-toolkit
-wrapt==1.17.2
-    # via deprecated
 zipp==3.23.0
    # via importlib-metadata
--- a/tools/autotag/components.xml
+++ b/tools/autotag/components.xml
@@ -1,7 +1,7 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <manifest>
    <remote name="rocm-org" fetch="https://github.com/ROCm/" />
-    <default revision="refs/tags/rocm-6.4.3"
+    <default revision="refs/tags/rocm-6.4.2"
     remote="rocm-org"
     sync-c="true"
     sync-j="4" />
--- a/tools/rocm-build/rocm-6.4.3.xml
+++ b/tools/rocm-build/rocm-6.4.3.xml
@@ -1,79 +0,0 @@
-<?xml version="1.0" encoding="UTF-8"?>
-<manifest>
-    <remote name="rocm-org" fetch="https://github.com/ROCm/" />
-    <default revision="refs/tags/rocm-6.4.3"
-     remote="rocm-org"
-     sync-c="true"
-     sync-j="4" />
-<!--list of projects for ROCm-->
-    <project name="ROCm" revision="roc-6.4.x" />
-    <project name="ROCK-Kernel-Driver" />
-    <project name="ROCR-Runtime" />
-    <project name="amdsmi" />
-    <project name="rdc" />
-    <project name="rocm_bandwidth_test" />
-    <project name="rocm_smi_lib" />
-    <project name="rocm-core" />
-    <project name="rocm-examples" />
-    <project name="rocminfo" />
-    <project name="rocprofiler" />
-    <project name="rocprofiler-register" />
-    <project name="rocprofiler-sdk" />
-    <project name="rocprofiler-compute" />
-    <project name="rocprofiler-systems" />
-    <project name="roctracer" />
-<!--HIP Projects-->
-    <project name="hip" />
-    <project name="hip-tests" />
-    <project name="HIPIFY" />
-    <project name="clr" />
-    <project name="hipother" />
-<!-- The following projects are all associated with the AMDGPU LLVM compiler -->
-    <project name="half" />
-    <project name="llvm-project" />
-    <project name="spirv-llvm-translator" />
-<!-- gdb projects -->
-    <project name="ROCdbgapi" />
-    <project name="ROCgdb" />
-    <project name="rocr_debug_agent" />
-<!-- ROCm Libraries -->
-    <project groups="mathlibs" name="AMDMIGraphX" />
-    <project groups="mathlibs" name="MIOpen" />
-    <project groups="mathlibs" name="MIVisionX" />
-    <project groups="mathlibs" name="ROCmValidationSuite" />
-    <project groups="mathlibs" name="Tensile" />
-    <project groups="mathlibs" name="composable_kernel" />
-    <project groups="mathlibs" name="hipBLAS-common" />
-    <project groups="mathlibs" name="hipBLAS" />
-    <project groups="mathlibs" name="hipBLASLt" />
-    <project groups="mathlibs" name="hipCUB" />
-    <project groups="mathlibs" name="hipFFT" />
-    <project groups="mathlibs" name="hipRAND" />
-    <project groups="mathlibs" name="hipSOLVER" />
-    <project groups="mathlibs" name="hipSPARSE" />
-    <project groups="mathlibs" name="hipSPARSELt" />
-    <project groups="mathlibs" name="hipTensor" />
-    <project groups="mathlibs" name="hipfort" />
-    <project groups="mathlibs" name="rccl" />
-    <project groups="mathlibs" name="rocAL" />
-    <project groups="mathlibs" name="rocALUTION" />
-    <project groups="mathlibs" name="rocBLAS" />
-    <project groups="mathlibs" name="rocDecode" />
-    <project groups="mathlibs" name="rocJPEG" />
-    <project groups="mathlibs" name="rocPyDecode" />
-    <project groups="mathlibs" name="rocFFT" />
-    <project groups="mathlibs" name="rocPRIM" />
-    <project groups="mathlibs" name="rocRAND" />
-    <project groups="mathlibs" name="rocSHMEM" />
-    <project groups="mathlibs" name="rocSOLVER" />
-    <project groups="mathlibs" name="rocSPARSE" />
-    <project groups="mathlibs" name="rocThrust" />
-    <project groups="mathlibs" name="rocWMMA" />
-    <project groups="mathlibs" name="rocm-cmake" />
-    <project groups="mathlibs" name="rpp" />
-    <project groups="mathlibs" name="TransferBench" />
-<!-- Projects for OpenMP-Extras -->
-    <project name="aomp" path="openmp-extras/aomp" />
-    <project name="aomp-extras" path="openmp-extras/aomp-extras" />
-    <project name="flang" path="openmp-extras/flang" />
-</manifest>
Author	SHA1	Message	Date
Matt Williams	b3211cc6fa	Updating broken link (#5258 )	2025-09-05 16:02:13 -04:00
Peter Park	5853468fca	Update PyTorch training benchmark docker doc to 25.7 (#5255 ) (#5260 ) * Update PyTorch training benchmark docker doc to 25.7 * update .wordlist.txt * update conf.py * update data sheet * fix sphinx warnings	2025-09-05 12:14:09 -04:00
Matt Williams	245c95690f	Merge pull request #5228 from ROCm/cherry-pick-link-fix Fix hyperlink syntax	2025-08-26 11:42:57 -04:00
Dominic Widdows	39c1b926f6	Fix hyperlink syntax	2025-08-26 11:35:30 -04:00
Jeffrey Novotny	3c3847f9f7	Merge pull request #5224 from amd-jnovotny/dlf-matt-docs643 Deep learning frameworks edits for scale (#5189)	2025-08-22 11:56:27 -04:00
Matt Williams	249bd177ec	Deep learning frameworks edits for scale (#5189 ) * Deep learning frameworks edits for scale Based on https://ontrack-internal.amd.com/browse/ROCDOC-1809 * update table table * leo comments * formatting * format * update table based on feedback * header * Update machine learning page * headers * Apply suggestions from code review Co-authored-by: anisha-amd <anisha.sankar@amd.com> * Update .wordlist.txt * formatting * Update docs/how-to/deep-learning-rocm.rst Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com> --------- Co-authored-by: Matt Williams <Matt.Williams+amdeng@amd.com> Co-authored-by: anisha-amd <anisha.sankar@amd.com> Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com> (cherry picked from commit `1d42f7cc62`)	2025-08-22 11:52:32 -04:00
Peter Park	b2ee8d4b2e	docs: Add Primus (Megatron) training Docker documentation (#5218 ) (#5222 ) (cherry picked from commit `98029db4ee`)	2025-08-22 09:02:40 -04:00
Peter Park	3f834cf520	Fix documented VRAM for Radeon AI Pro R9700 (#5203 ) (#5204 ) (cherry picked from commit `c154b7e0a3`)	2025-08-18 10:20:22 -04:00
Peter Park	70ba866c5b	vLLM inference benchmark doc: add missing data field (#5199 ) (#5200 ) (cherry picked from commit `55d0a88ec5`)	2025-08-15 13:52:51 -04:00
Peter Park	320ec4669a	[docs/6.4.3] Update vLLM benchmark doc for 20250812 Docker release (#5196 ) (#5198 ) (cherry picked from commit `7ee22790ce`)	2025-08-14 15:51:58 -04:00
anisha-amd	c9bd93b537	[Docs] 6.4.3: compatibility matrix frameworks support update (#5186 )	2025-08-12 14:25:45 -04:00
Peter Park	a060550bcd	Add Hunyuan Video to PyTorch inference benchmark models doc (#5094 ) (cherry picked from commit `80f7dc79b9`)	2025-08-12 11:59:51 -04:00
Parag Bhandari	c92cbaee66	Merge branch 'roc-6.4.x' into docs/6.4.3	2025-08-08 08:49:53 -04:00
Parag Bhandari	c84afacc8d	Merge branch 'develop' into roc-6.4.x	2025-08-08 08:49:39 -04:00
Parag Bhandari	843fd1b3fb	Merge branch 'roc-6.4.x' into docs/6.4.3	2025-08-08 07:21:31 -04:00
Parag Bhandari	82221c4e2d	Merge branch 'develop' into roc-6.4.x	2025-08-08 07:18:53 -04:00
pbhandar-amd	d0ebe126e7	Sync develop into docs/6.4.3	2025-08-07 09:07:38 -04:00
pbhandar-amd	74610893a9	Merge pull request #5154 from ROCm/amd/pbhandar/merge_from_develop Merge develop into docs/6.4.3	2025-08-06 11:25:50 -04:00
Parag Bhandari	afe3e21cad	Merge branch 'develop' into docs/6.4.3	2025-08-05 16:24:07 -04:00
Alex Xu	ae2440772f	upgrade rocm-docs-core to 1.22.0	2025-07-31 16:41:40 -04:00
anisha-amd	61f970a24d	Cherry pick into 6.4.3 - Docs: Adding frameworks compatibility for Megablocks and Taichi (#5138 )	2025-07-31 14:02:09 -04:00
Alex Xu	85a1682573	Merge branch 'develop' into roc-6.4.x	2025-07-21 17:22:49 -04:00
Alex Xu	87c6e320b4	Merge branch 'develop' into roc-6.4.x	2025-07-21 15:52:30 -04:00
ammallya	b50948fe6b	Fix for rocrsamples and rocr_debug_agent (#4863 ) * Fix for rocrsamples * Fix for rocr_debug_agent	2025-05-30 16:27:29 -07:00
ammallya	91407405a9	Changed naming convention for hip (#4837 ) * Changed naming convention for hip * Changed naming convention for hip	2025-05-29 10:19:28 -07:00
ammallya	8f23f63a6b	Fix for tests (#4818 ) * Fix for RBT * Fix for roctst and kfd test	2025-05-27 17:38:48 -07:00
Alex Xu	11747aaadc	Merge branch 'develop' into roc-6.4.x	2025-05-21 15:04:02 -04:00
Alex Xu	1088beefe5	Merge branch 'develop' into roc-6.4.x	2025-05-21 12:27:13 -04:00
Alex Xu	b7988925a5	Merge branch 'develop' into roc-6.4.x	2025-05-21 12:25:30 -04:00
chiranjeevipattigidi	89dafa6232	Update packages - remove broken packages (#4758 ) * Update envsetup.sh HIP_ON_ROCclr_ROOT path to hip and remove aqlprofiletest * Update packages - remove broken packages	2025-05-21 09:06:39 -07:00
chiranjeevipattigidi	8054852dad	Update envsetup.sh HIP_ON_ROCclr_ROOT path to hip and remove (#4755 ) aqlprofiletest	2025-05-20 07:59:07 -07:00
ammallya	542d7813ce	Removing aqlprofiletest	2025-04-14 15:26:24 -07:00
ammallya	bc1ffe4fcb	bypass tests	2025-04-14 13:41:34 -07:00
ammallya	09997c68bb	Removing kfd test	2025-04-14 12:55:13 -07:00
ammallya	42bc3501ac	Merge pull request #4623 from ammallya/roc-6.4.x Rebasing branch 6.4.x	2025-04-14 11:42:06 -07:00