[jira] [Assigned] (SPARK-31976) use MemoryUsage to control the size of block
[ https://issues.apache.org/jira/browse/SPARK-31976?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] zhengruifeng reassigned SPARK-31976: Assignee: zhengruifeng > use MemoryUsage to control the size of block > > > Key: SPARK-31976 > URL: https://issues.apache.org/jira/browse/SPARK-31976 > Project: Spark > Issue Type: Sub-task > Components: ML, PySpark >Affects Versions: 3.1.0 >Reporter: zhengruifeng >Assignee: zhengruifeng >Priority: Major > > According to the performance test in > https://issues.apache.org/jira/browse/SPARK-31783, the performance gain is > mainly related to the nnz of block. > So it maybe reasonable to control the size of block by memory usage, instead > of number of rows. > > note1: param blockSize had already used in ALS and MLP to stack vectors > (expected to be dense); > note2: we may refer to the {{Strategy.maxMemoryInMB}} in tree models; > > There may be two ways to impl: > 1, compute the sparsity of input vectors ahead of train (this can be computed > with other statistics computation, maybe no extra pass), and infer a > reasonable number of vectors to stack; > 2, stack the input vectors adaptively, by monitoring the memory usage in a > block; -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Assigned] (SPARK-31976) use MemoryUsage to control the size of block
[ https://issues.apache.org/jira/browse/SPARK-31976?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Apache Spark reassigned SPARK-31976: Assignee: Apache Spark > use MemoryUsage to control the size of block > > > Key: SPARK-31976 > URL: https://issues.apache.org/jira/browse/SPARK-31976 > Project: Spark > Issue Type: Sub-task > Components: ML, PySpark >Affects Versions: 3.1.0 >Reporter: zhengruifeng >Assignee: Apache Spark >Priority: Major > > According to the performance test in > https://issues.apache.org/jira/browse/SPARK-31783, the performance gain is > mainly related to the nnz of block. > So it maybe reasonable to control the size of block by memory usage, instead > of number of rows. > > note1: param blockSize had already used in ALS and MLP to stack vectors > (expected to be dense); > note2: we may refer to the {{Strategy.maxMemoryInMB}} in tree models; > > There may be two ways to impl: > 1, compute the sparsity of input vectors ahead of train (this can be computed > with other statistics computation, maybe no extra pass), and infer a > reasonable number of vectors to stack; > 2, stack the input vectors adaptively, by monitoring the memory usage in a > block; -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Assigned] (SPARK-31976) use MemoryUsage to control the size of block
[ https://issues.apache.org/jira/browse/SPARK-31976?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Apache Spark reassigned SPARK-31976: Assignee: (was: Apache Spark) > use MemoryUsage to control the size of block > > > Key: SPARK-31976 > URL: https://issues.apache.org/jira/browse/SPARK-31976 > Project: Spark > Issue Type: Sub-task > Components: ML, PySpark >Affects Versions: 3.1.0 >Reporter: zhengruifeng >Priority: Major > > According to the performance test in > https://issues.apache.org/jira/browse/SPARK-31783, the performance gain is > mainly related to the nnz of block. > So it maybe reasonable to control the size of block by memory usage, instead > of number of rows. > > note1: param blockSize had already used in ALS and MLP to stack vectors > (expected to be dense); > note2: we may refer to the {{Strategy.maxMemoryInMB}} in tree models; > > There may be two ways to impl: > 1, compute the sparsity of input vectors ahead of train (this can be computed > with other statistics computation, maybe no extra pass), and infer a > reasonable number of vectors to stack; > 2, stack the input vectors adaptively, by monitoring the memory usage in a > block; -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org