output: /databricks/spark/sql/mllib-perf-ci
timeoutSeconds: 43200 # This limit is for all benchmarks and should be bumped as more are added.
common:
  numExamples: 1000000
  numTestExamples: 1000000
  numFeatures: 4000
  numPartitions: 64
  randomSeed: [1, 1, 1] # Rerun 3 times to accumulate some info
benchmarks:
  - name: classification.DecisionTreeClassification
    params:
      depth: [5, 12]
      numClasses: 4
  - name: classification.GBTClassification
    params:
      numFeatures: 1500
      depth: 4
      numClasses: 4
      maxIter: 8
  - name: classification.RandomForestClassification
    params:
      depth: 10
      numFeatures: 1000
      numClasses: 4
      maxIter: 100
  - name: classification.LogisticRegression
    params:
      numExamples: 700000
      numFeatures: 5000
      elasticNetParam: [0.0, 0.5]
      regParam: 0.01
      tol: 0.0
      maxIter: 10
  - name: classification.LinearSVC
    params:
      numExamples: 500000
      regParam: 0.01
      tol: 0
      maxIter: 10
  - name: classification.NaiveBayes
    params:
      numExamples: 2000000
      numFeatures: 5000
      numClasses: 10
      smoothing: 1.0
  - name: clustering.GaussianMixture
    params:
      numFeatures: 30
      k: 15
      maxIter: 15
      tol: 0.0
  - name: clustering.KMeans
    params:
      k: 20
      maxIter: 20
      tol: 1e-3
  - name: clustering.LDA
    params:
      numExamples: 200000
      docLength: 100
      vocabSize: 5000
      k: 20
      maxIter: 10
      optimizer:
        - em
        - online
  - name: feature.BucketedRandomProjectionLSH
    params:
      numHashTables: 20
  - name: feature.Bucketizer
    params:
      bucketizerNumBuckets: 20
  - name: feature.HashingTF
    params:
      docLength: 2000
      vocabSize: 200
  - name: feature.MinHashLSH
    params:
      numFeatures: 10000
      numHashTables: 10
  - name: feature.OneHotEncoder
    params:
      featureArity: 10000
  - name: feature.StringIndexer
    params:
      vocabSize: 10000
  - name: feature.Tokenizer
    params:
      vocabSize: 2000
      docLength: 10000
  - name: feature.VectorAssembler
    params:
      numInputCols: 20
  - name: feature.VectorSlicer
    params:
      numFeatures: 5000
  - name: feature.Word2Vec
    params:
      numExamples: 10000
      vocabSize: 100
      docLength: 300
      numSynonymsToFind: 3
  - name: fpm.FPGrowth
    params:
      numExamples: 10000000
      numItems: 10000
      itemSetSize: [4, 14]
  - name: recommendation.ALS
    params:
      numExamples: 20000000
      numTestExamples: 50000000
      numUsers: 4000000
      numItems: 4000000
      regParam: 0.01
      rank: 10
      maxIter: 8
  - name: regression.DecisionTreeRegression
    params:
      depth: [4, 7]
  - name: regression.GBTRegression
    params:
      numFeatures: 1000
      depth: 4
      maxIter: 8
  - name: regression.GLMRegression
    params:
      numExamples: 400000
      numTestExamples: 500000
      numFeatures: 500
      link: log
      family: gaussian
      tol: 0.0
      maxIter: 8
      regParam: 0.1
  - name: regression.LinearRegression
    params:
      numFeatures: 5000
      regParam: 0.01
      tol: 0.0
      maxIter: 9
  - name: regression.RandomForestRegression
    params:
      depth: 7
      numFeatures: 500
      maxIter: 16