Commits on Aug 6, 2024

1. Support int64 index dtype in sparse logistic regression (#700) …

   * in progress
   
   * setup large csr generation and testing with LR
   
   * revise core.py to avoid using cupyx when nnz of a partition is larger than int32 max
   
   * in progress get uvm work with cupyx int32 limit
   
   * support testing int64 with large csr_matrix/sparse vectors
   
   * support testing on large dataset with new session
   
   * resolve cleansparksession must aligns confs
   
   * optimize with a comment
   
   * increase memory and parallelism for testing large dataset
   
   * fix a kmeans test failure that is due to latest cuml introduces a new parameter to sg kmeans
   
   * update ci to cuml 24.08 nightly
   
   * revise test case to include sparsedatagen
   
   * remove nit
   
   * revised driver.meomry to 32g, let cache on disk
   
   * replace row-wise udf with mapInPandas
   
   * use pythonpath instead of os.insert to include benchmark folder
   
   ---------
   
   Signed-off-by: Jinfeng <[email protected]>

   

   lijinf2 committed Aug 6, 2024
   Configuration menu

   • View commit details
   • Copy full SHA for bf4554f
   • Browse repository at this point

   Copy the full SHA

   bf4554f View commit details

   Browse the repository at this point in the history