Bitpack Huffman tables (#786)#786
Open
terrelln wants to merge 9 commits into
Open
Conversation
Differential Revision: D106089170
Differential Revision: D106089168
Summary: LZ must be a dynamic graph because it invokes a multi-input node (ZL_NODE_MUX_LENGTHS). But we need to allow overriding the successors in a serialized graph for training and configurability. So do the same thing that FieldLZ does, which sets the index of the `customGraph` as a local int param. When the local int param is set, set the corresponding successor to that graph. Differential Revision: D105367885
Summary: As title Differential Revision: D105368972
Differential Revision: D102620101
Differential Revision: D102351635
Differential Revision: D102352000
|
@terrelln has exported this pull request. If you are a Meta employee, you can view the originating Diff in D105622794. |
Summary: Pull Request resolved: facebook#786 This simplification is a major speed win on small data for a small header size regression. On large data, which is the current focus of OpenZL, the difference is negligible, for both speed and ratio. A future diff will introduce a new header encoding codec that will achieve the speed of bitpack with the ratio of FSE. I have a draft working, but still have to clean it up. In the meantime we will bitpack. Reviewed By: Cyan4973 Differential Revision: D105622794
meta-codesync
Bot
changed the title
terrelln
force-pushed
the
export-D105622794
branch
from
292934d to
48c8b9f
Compare
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary:
This simplification is a major speed win on small data for a small header size regression. On large data, which is the current focus of OpenZL, the difference is negligible, for both speed and ratio.
A future diff will introduce a new header encoding codec that will achieve the speed of bitpack with the ratio of FSE. I have a draft working, but still have to clean it up. In the meantime we will bitpack.
Reviewed By: Cyan4973
Differential Revision: D105622794