fall back to cudf22.04 for runtime error

See issue: NVIDIA#73

Signed-off-by: Allen Xu <[email protected]>

README.md

@@ -49,7 +49,7 @@ and cmake dependecies
 4. [cuDF](https://github.com/rapidsai/cudf):
     - install cuDF shared library via conda:
       ```bash
-      conda install -c rapidsai-nightly -c nvidia -c conda-forge cudf=22.06 python=3.8 -y
+      conda install -c rapidsai-nightly -c nvidia -c conda-forge cudf=22.04 python=3.8 -y
       ```
 5. [RAFT(22.06)](https://github.com/rapidsai/raft):
     - raft provides only header files, so no build instructions for it.
@@ -101,4 +101,4 @@ more details about example code. We provide both
 [Notebook](https://github.com/NVIDIA/spark-rapids-examples/blob/branch-22.06/examples/Spark-cuML/pca/PCA-example-notebook.ipynb)
 and [jar](https://github.com/NVIDIA/spark-rapids-examples/blob/branch-22.06/examples/Spark-cuML/pca/scala/src/com/nvidia/spark/examples/pca/Main.scala)
  versions there. Instructions to run these examples are described in the
- [README](https://github.com/NVIDIA/spark-rapids-examples/blob/branch-22.06/examples/Spark-cuML/pca/README.md).
+ [README](https://github.com/NVIDIA/spark-rapids-examples/blob/branch-22.06/examples/Spark-cuML/pca/README.md).

native/src/CMakeLists.txt

@@ -59,9 +59,9 @@ include(${CPM_DOWNLOAD_LOCATION})
 # pull cuDF sources, to use jni_utils.hpp
 # cmake options should be added here for CI build.
 CPMAddPackage(NAME cudf
-        VERSION         "22.06.00"
+        VERSION         "22.04.00"
         GIT_REPOSITORY  https://github.com/rapidsai/cudf.git
-        GIT_TAG         branch-22.06
+        GIT_TAG         branch-22.04
 )
 
 add_library(rapidsml_jni SHARED rapidsml_jni.cpp