[SPARK-12685] word2vec trainWordsCount gets overflow

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 1.6.0
• Fix Version/s: 1.4.2, 1.5.3, 1.6.1, 2.0.0
• Component/s: MLlib
• Target Version/s: 1.4.2, 1.5.3, 1.6.1, 2.0.0
• Assignee: yuhao yang
• Reporter: yuhao yang
• Shepherd: Joseph K. Bradley
• Created: 07/Jan/16 04:28
• Updated: 13/Jan/16 19:54
• Resolved: 13/Jan/16 19:54

Description

the log of word2vec reports
trainWordsCount = -785727483
during computation over a large dataset.

I'll also add vocabsize to the log.

Update the priority as it will affects the computation process.
alpha =
learningRate * (1 - numPartitions * wordCount.toDouble / (trainWordsCount + 1))

Issue Links

• links to: [Github] Pull Request #10627 (hhbyyh)
• links to: [Github] Pull Request #10721 (hhbyyh)