feat: Add VowpalWabbit ngram support #696
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
/azp run |
|
Azure Pipelines successfully started running 1 pipeline(s). |
eisber
changed the title
Codecov Report
@@ Coverage Diff @@
## master #696 +/- ##
==========================================
+ Coverage 79.99% 80.03% +0.04%
==========================================
Files 229 229
Lines 9118 9142 +24
Branches 469 470 +1
==========================================
+ Hits 7294 7317 +23
- Misses 1824 1825 +1
Continue to review full report at Codecov.
|
eisber
changed the title
|
/azp run |
|
Azure Pipelines successfully started running 1 pipeline(s). |
|
/azp run |
|
Azure Pipelines successfully started running 1 pipeline(s). |
|
/azp run |
|
Azure Pipelines successfully started running 1 pipeline(s). |
mhamilton723
previously approved these changes
Oct 8, 2019
src/main/scala/com/microsoft/ml/spark/vw/VowpalWabbitFeaturizer.scala
Outdated
Show resolved
Hide resolved
src/test/scala/com/microsoft/ml/spark/vw/VerifyVowpalWabbitClassifier.scala
Outdated
Show resolved
Hide resolved
|
/azp run |
|
Azure Pipelines successfully started running 1 pipeline(s). |
|
/azp run |
|
Azure Pipelines successfully started running 1 pipeline(s). |
|
/azp run |
|
Azure Pipelines successfully started running 1 pipeline(s). |
mhamilton723
approved these changes
Oct 9, 2019
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Vowpal Wabbit supports on-the-fly ngram generation, thus the featurized dataset does not need to contain the ngrams and will not blow up in memory. Minor complication in all of this is that the features are already hashed in the SparseVector and unfortunately always get sorted by the Spark SparseVector implementation.
To overcome this the sparse indices get prefixed with a counter at the highest level bits. The number of bits for the counter can be configured, but together with the hashing bits need to fit within 30 bits (yeah, signed integers). The prefix counter is stripped by VW JNI layer as the bit mask is always re-applied.