llm-ppl-evaluation.yml

name: LLM Perplexity Evaluation

# Cancel previous runs in the PR when you push new commits
concurrency:
  group: ${{ github.workflow }}-llm-nightly-test-${{ github.event.pull_request.number || github.run_id }}
  cancel-in-progress: true

permissions:
  contents: read

# Controls when the action will run.
on:
  schedule:
    - cron: "00 14 * * *" # GMT time, 14:00 GMT == 22:00 China
  pull_request:
    branches: [main]
    paths:
      - ".github/workflows/llm-ppl-evaluation.yml"
  # Allows you to run this workflow manually from the Actions tab
  workflow_dispatch:
    inputs:
      seq_len:
        description: 'sequence length, input one number and must be quoted.'
        required: true
        type: string      
      model_name:
        description: 'Model names, separated by comma and must be quoted.'
        required: true
        type: string
      precision:
        description: 'Precisions, separated by comma and must be quoted.'
        required: true
        type: string
      language:
        description: 'language, can be en, zh, or all and must be quoted.'
        required: true
        type: string
      runs-on:
        description: 'Labels to filter the runners, separated by comma and must be quoted.'
        default: "accuracy"
        required: false
        type: string


# A workflow run is made up of one or more jobs that can run sequentially or in parallel
jobs:
  llm-cpp-build:
    uses: ./.github/workflows/llm-binary-build.yml
  set-matrix:
    runs-on: ubuntu-latest
    outputs:
      seq_len: ${{ steps.set-matrix.outputs.seq_len }}
      model_name: ${{ steps.set-matrix.outputs.model_name }}
      precision: ${{ steps.set-matrix.outputs.precision }}
      language: ${{ steps.set-matrix.outputs.language }}
      runner: ${{ steps.set-matrix.outputs.runner }}
    steps:
      - name: set-nightly-env
        if: ${{github.event_name == 'schedule'}}
        env:
          NIGHTLY_MATRIX_SEQ_LEN: '["512"]'
          NIGHTLY_MATRIX_MODEL_NAME: '["Llama-2-7b-chat-hf", "mpt-7b-chat", 
                        "falcon-7b-instruct-with-patch", "Mistral-7B-v0.1"]'
          NIGHTLY_MATRIX_LANGUAGE: '["en"]'
          NIGHTLY_MATRIX_PRECISION: '["sym_int4 fp8"]'
          NIGHTLY_LABELS: '["self-hosted", "llm", "accuracy-nightly"]'
        run: |
            echo "seq_len=$NIGHTLY_MATRIX_SEQ_LEN" >> $GITHUB_ENV
            echo "model_name=$NIGHTLY_MATRIX_MODEL_NAME" >> $GITHUB_ENV
            echo "precision=$NIGHTLY_MATRIX_PRECISION" >> $GITHUB_ENV
            echo "language=$NIGHTLY_MATRIX_LANGUAGE" >> $GITHUB_ENV
            echo "runner=$NIGHTLY_LABELS" >> $GITHUB_ENV

      - name: set-pr-env
        if: ${{github.event_name == 'pull_request'}}
        env:
          PR_MATRIX_SEQ_LEN: '["512"]'
          PR_MATRIX_MODEL_NAME: '["Llama-2-7b-chat-hf"]'
          PR_MATRIX_LANGUAGE: '["en"]'
          PR_MATRIX_PRECISION: '["sym_int4"]'
          PR_LABELS: '["self-hosted", "llm", "temp-arc01"]'
        run: |
            echo "seq_len=$PR_MATRIX_SEQ_LEN" >> $GITHUB_ENV
            echo "model_name=$PR_MATRIX_MODEL_NAME" >> $GITHUB_ENV
            echo "precision=$PR_MATRIX_PRECISION" >> $GITHUB_ENV
            echo "language=$PR_MATRIX_LANGUAGE" >> $GITHUB_ENV
            echo "runner=$PR_LABELS" >> $GITHUB_ENV
      - name: set-manual-env
        if: ${{github.event_name == 'workflow_dispatch'}}
        env:
          MANUAL_MATRIX_SEQ_LEN: ${{format('[ {0} ]', inputs.seq_len)}}
          MANUAL_MATRIX_MODEL_NAME: ${{format('[ {0} ]', inputs.model_name)}}
          MANUAL_MATRIX_LANGUAGE: ${{format('[ {0} ]', inputs.language)}}
          MANUAL_MATRIX_PRECISION: ${{format('[ {0} ]', inputs.precision)}}
          MANUAL_LABELS: ${{format('["self-hosted", "llm", {0}]', inputs.runs-on)}}
        run: |
            echo "seq_len=$MANUAL_MATRIX_SEQ_LEN" >> $GITHUB_ENV
            echo "model_name=$MANUAL_MATRIX_MODEL_NAME" >> $GITHUB_ENV
            echo "precision=$MANUAL_MATRIX_PRECISION" >> $GITHUB_ENV
            echo "language=$MANUAL_MATRIX_LANGUAGE" >> $GITHUB_ENV
            echo "runner=$MANUAL_LABELS" >> $GITHUB_ENV
      - name: set-matrix
        id: set-matrix
        run: |
            echo "seq_len=$seq_len" >> $GITHUB_OUTPUT
            echo "model_name=$model_name" >> $GITHUB_OUTPUT
            echo "precision=$precision" >> $GITHUB_OUTPUT
            echo "language=$language" >> $GITHUB_OUTPUT
            echo "runner=$runner" >> $GITHUB_OUTPUT
  llm-ppl-evaluation:
    timeout-minutes: 1000
    needs: [llm-cpp-build, set-matrix]
    strategy:
      fail-fast: false
      matrix:
        # include:
        #   python-version: "3.9"
        #   model_name: "stablelm-3b-4e1t"
        #   task: "arc"
        #   precision: "sym_int4" #options: sym_int4, fp4, mixed_fp4, sym_int8, fp8, mixed_fp8
        python-version: ["3.9"]
        model_name: ${{ fromJson(needs.set-matrix.outputs.model_name) }}
        language: ${{ fromJson(needs.set-matrix.outputs.language) }}
        precision: ${{ fromJson(needs.set-matrix.outputs.precision) }}
        seq_len: ${{ fromJson(needs.set-matrix.outputs.seq_len) }}
        device: [xpu]
        include:
          - python-version: "3.9"
            model_name: "chatglm2-6b"
            language: "zh"
            precision: "sym_int4 fp8"
            seq_len: "512"
            device: "xpu"
          - python-version: "3.9"
            model_name: "chatglm3-6b"
            language: "zh"
            precision: "sym_int4 fp8"
            seq_len: "512"
            device: "xpu"
          - python-version: "3.9"
            model_name: "Baichuan2-7B-Chat"
            language: "zh"
            precision: "sym_int4 fp8"
            seq_len: "512"
            device: "xpu"
        
    runs-on: ${{ fromJson(needs.set-matrix.outputs.runner) }}
    env:
      ANALYTICS_ZOO_ROOT: ${{ github.workspace }}
      ORIGIN_DIR: /mnt/disk1/models
      DATASET_DIR: /mnt/disk1/datasets/THUDM___long_bench/
    steps:
      - uses: actions/checkout@f43a0e5ff2bd294095638e18286ca9a3d1956744 # actions/checkout@v3
      - name: Set up Python ${{ matrix.python-version }}
        uses: actions/setup-python@v4
        with:
          python-version: ${{ matrix.python-version }}
      - name: Install dependencies
        shell: bash
        run: |
          python -m pip install --upgrade pip
          python -m pip install --upgrade setuptools==58.0.4
          python -m pip install --upgrade wheel

      - name: Download llm binary
        uses: ./.github/actions/llm/download-llm-binary

      - name: Run LLM install (all) test
        uses: ./.github/actions/llm/setup-llm-env
        with:
          extra-dependency: "xpu_2.1"
      
      - name: Download models
        shell: bash
        run: |
          echo "MODEL_PATH=${ORIGIN_DIR}/${{ matrix.model_name }}/" >> "$GITHUB_ENV"
          MODEL_PATH=${ORIGIN_DIR}/${{ matrix.model_name }}/
          wget -r -nH -nc --no-verbose --cut-dirs=1 ${LLM_FTP_URL}/llm/${{ matrix.model_name }} -P ${ORIGIN_DIR}
          
      - name: Upgrade packages
        shell: bash
        run: |
          pip install --upgrade datasets==2.14.6 
          if [ "${{ matrix.model_name }}" = "Mistral-7B-v0.1" ]; then
          pip install --upgrade transformers==4.36
          else
          pip install --upgrade transformers==4.31
          fi

      - name: Run perplexity
        shell: bash
        working-directory: ${{ github.workspace }}/python/llm/dev/benchmark/perplexity
        env:
          USE_XETLA: OFF
          SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS: 1
        run: |
          source /opt/intel/oneapi/setvars.sh
          python run.py \
          --seq_len ${{ matrix.seq_len }} \
          --model_path ${MODEL_PATH} \
          --precisions ${{ matrix.precision }} \
          --device ${{ matrix.device }} \
          --dataset_path ${DATASET_DIR} \
          --language ${{ matrix.language }}