Upgrade to CUDA 10.0 (dmlc#5649)

Co-authored-by: fis <[email protected]>

Author: hcho3

Jenkinsfile

@@ -64,7 +64,6 @@ pipeline {
             'build-cpu': { BuildCPU() },
             'build-cpu-rabit-mock': { BuildCPUMock() },
             'build-cpu-non-omp': { BuildCPUNonOmp() },
-            'build-gpu-cuda9.0': { BuildCUDA(cuda_version: '9.0') },
             'build-gpu-cuda10.0': { BuildCUDA(cuda_version: '10.0') },
             'build-gpu-cuda10.1': { BuildCUDA(cuda_version: '10.1') },
             'build-jvm-packages': { BuildJVMPackages(spark_version: '2.4.3') },
@@ -251,10 +250,10 @@ def BuildCUDA(args) {
     ${dockerRun} ${container_type} ${docker_binary} ${docker_args} bash -c "cd python-package && rm -rf dist/* && python setup.py bdist_wheel --universal"
     ${dockerRun} ${container_type} ${docker_binary} ${docker_args} python3 tests/ci_build/rename_whl.py python-package/dist/*.whl ${commit_id} manylinux2010_x86_64
     """
-    // Stash wheel for CUDA 9.0 target
-    if (args.cuda_version == '9.0') {
+    // Stash wheel for CUDA 10.0 target
+    if (args.cuda_version == '10.0') {
       echo 'Stashing Python wheel...'
-      stash name: 'xgboost_whl_cuda9', includes: 'python-package/dist/*.whl'
+      stash name: 'xgboost_whl_cuda10', includes: 'python-package/dist/*.whl'
       path = ("${BRANCH_NAME}" == 'master') ? '' : "${BRANCH_NAME}/"
       s3Upload bucket: 'xgboost-nightly-builds', path: path, acl: 'PublicRead', workingDir: 'python-package/dist', includePathPattern:'**/*.whl'
       echo 'Stashing C++ test executable (testxgboost)...'
@@ -298,7 +297,7 @@ def BuildJVMDoc() {
 
 def TestPythonCPU() {
   node('linux && cpu') {
-    unstash name: 'xgboost_whl_cuda9'
+    unstash name: 'xgboost_whl_cuda10'
     unstash name: 'srcs'
     unstash name: 'xgboost_cli'
     echo "Test Python CPU"
@@ -315,7 +314,7 @@ def TestPythonCPU() {
 def TestPythonGPU(args) {
   nodeReq = (args.multi_gpu) ? 'linux && mgpu' : 'linux && gpu'
   node(nodeReq) {
-    unstash name: 'xgboost_whl_cuda9'
+    unstash name: 'xgboost_whl_cuda10'
     unstash name: 'srcs'
     echo "Test Python GPU: CUDA ${args.cuda_version}"
     def container_type = "gpu"

tests/python-gpu/test_gpu_basic_models.py

@@ -12,25 +12,15 @@
 class TestGPUBasicModels(unittest.TestCase):
     cputest = test_bm.TestModels()
 
-    def test_eta_decay_gpu_hist(self):
-        self.cputest.run_eta_decay('gpu_hist')
-
-    def test_deterministic_gpu_hist(self):
-        kRows = 1000
-        kCols = 64
-        kClasses = 4
-        # Create large values to force rounding.
-        X = np.random.randn(kRows, kCols) * 1e4
-        y = np.random.randint(0, kClasses, size=kRows)
-
+    def run_cls(self, X, y, deterministic):
         cls = xgb.XGBClassifier(tree_method='gpu_hist',
-                                deterministic_histogram=True,
+                                deterministic_histogram=deterministic,
                                 single_precision_histogram=True)
         cls.fit(X, y)
         cls.get_booster().save_model('test_deterministic_gpu_hist-0.json')
 
         cls = xgb.XGBClassifier(tree_method='gpu_hist',
-                                deterministic_histogram=True,
+                                deterministic_histogram=deterministic,
                                 single_precision_histogram=True)
         cls.fit(X, y)
         cls.get_booster().save_model('test_deterministic_gpu_hist-1.json')
@@ -40,7 +30,24 @@ def test_deterministic_gpu_hist(self):
         with open('test_deterministic_gpu_hist-1.json', 'r') as fd:
             model_1 = fd.read()
 
-        assert hash(model_0) == hash(model_1)
-
         os.remove('test_deterministic_gpu_hist-0.json')
         os.remove('test_deterministic_gpu_hist-1.json')
+
+        return hash(model_0), hash(model_1)
+
+    def test_eta_decay_gpu_hist(self):
+        self.cputest.run_eta_decay('gpu_hist')
+
+    def test_deterministic_gpu_hist(self):
+        kRows = 1000
+        kCols = 64
+        kClasses = 4
+        # Create large values to force rounding.
+        X = np.random.randn(kRows, kCols) * 1e4
+        y = np.random.randint(0, kClasses, size=kRows) * 1e4
+
+        model_0, model_1 = self.run_cls(X, y, True)
+        assert model_0 == model_1
+
+        model_0, model_1 = self.run_cls(X, y, False)
+        assert model_0 != model_1