update

minmingzhu · Nov 1, 2024 · 1d8521d · 1d8521d
1 parent 06d7caa
commit 1d8521d
Show file tree

Hide file tree

Showing 4 changed files with 14 additions and 20 deletions.
diff --git a/mllib-dal/src/main/native/DecisionForestClassifierImpl.cpp b/mllib-dal/src/main/native/DecisionForestClassifierImpl.cpp
@@ -286,11 +286,6 @@ static jobject doRFClassifierOneAPICompute(
         preview::train(comm, df_desc, hFeaturetable, hLabeltable);
     const auto result_infer =
         preview::infer(comm, df_desc, result_train.get_model(), hFeaturetable);
-    t2 = std::chrono::high_resolution_clock::now();
-    duration =
-        (float)std::chrono::duration_cast<std::chrono::milliseconds>(t2 -
-                                                                     t1)
-            .count();
     jobject trees = nullptr;
     if (isRoot) {
         logger::println(logger::INFO, "Variable importance results:");
@@ -300,7 +295,8 @@ static jobject doRFClassifierOneAPICompute(
         logger::println(logger::INFO, "Prediction results:");
         printHomegenTable(result_infer.get_responses());
         logger::println(logger::INFO, "Probabilities results:\n");
-        printHomegenTable(result_infer.get_probabilities());
+        printHomegenTable(result_infer.get_probabilities());ll
+
         t2 = std::chrono::high_resolution_clock::now();
         duration =
             (float)std::chrono::duration_cast<std::chrono::milliseconds>(t2 -

diff --git a/mllib-dal/src/main/native/DecisionForestRegressorImpl.cpp b/mllib-dal/src/main/native/DecisionForestRegressorImpl.cpp
@@ -276,11 +276,6 @@ static jobject doRFRegressorOneAPICompute(
         preview::train(comm, df_desc, hFeaturetable, hLabeltable);
     const auto result_infer =
         preview::infer(comm, df_desc, result_train.get_model(), hFeaturetable);
-    t2 = std::chrono::high_resolution_clock::now();
-    duration =
-        (float)std::chrono::duration_cast<std::chrono::milliseconds>(t2 -
-                                                                     t1)
-            .count();
     jobject trees = nullptr;
     if (isRoot) {
         logger::println(logger::INFO, "Variable importance results:");
@@ -289,7 +284,6 @@ static jobject doRFRegressorOneAPICompute(
         printHomegenTable(result_train.get_oob_err());
         logger::println(logger::INFO, "Prediction results:");
         printHomegenTable(result_infer.get_responses());
-
         t2 = std::chrono::high_resolution_clock::now();
         duration =
             (float)std::chrono::duration_cast<std::chrono::milliseconds>(t2 -

diff --git a/...dal/src/main/scala/com/intel/oap/mllib/classification/RandomForestClassifierDALImpl.scala b/...dal/src/main/scala/com/intel/oap/mllib/classification/RandomForestClassifierDALImpl.scala
@@ -79,10 +79,12 @@ class RandomForestClassifierDALImpl(val uid: String,
     val kvsIPPort = getOneCCLIPPort(labeledPointsTables)
     val training_breakdown_name = "RFClassifier_training_breakdown_" + executorNum;
 
-    labeledPointsTables.mapPartitionsWithIndex { (rank, table) =>
-      OneCCL.init(executorNum, rank, kvsIPPort, training_breakdown_name, storePath)
-      Iterator.empty
-    }.count()
+    if (useDevice == "CPU") {
+        labeledPointsTables.mapPartitionsWithIndex { (rank, table) =>
+          OneCCL.init(executorNum, rank, kvsIPPort, training_breakdown_name, storePath)
+          Iterator.empty
+        }.count()
+    }
     rfcTimer.record("OneCCL Init")
 
     val results = labeledPointsTables.mapPartitionsWithIndex { (rank, tables) =>

diff --git a/mllib-dal/src/main/scala/com/intel/oap/mllib/regression/RandomForestRegressorDALImpl.scala b/mllib-dal/src/main/scala/com/intel/oap/mllib/regression/RandomForestRegressorDALImpl.scala
@@ -73,10 +73,12 @@ class RandomForestRegressorDALImpl(val uid: String,
     val kvsIPPort = getOneCCLIPPort(labeledPointsTables)
     val training_breakdown_name = "RFRegressor_training_breakdown_" + executorNum;
 
-    labeledPointsTables.mapPartitionsWithIndex { (rank, table) =>
-      OneCCL.init(executorNum, rank, kvsIPPort, training_breakdown_name, storePath)
-      Iterator.empty
-    }.count()
+    if (useDevice == "CPU") {
+        labeledPointsTables.mapPartitionsWithIndex { (rank, table) =>
+          OneCCL.init(executorNum, rank, kvsIPPort, training_breakdown_name, storePath)
+          Iterator.empty
+        }.count()
+    }
     rfrTimer.record("OneCCL Init")
 
     val results = labeledPointsTables.mapPartitionsWithIndex { (rank, tables) =>