Add support for remapping node prediction results. #577
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
Originally, we save prediction results only. However, unlink node prediction tasks, it is impossible to find the corresponding edges for the edge prediction results. In this PR, other than storing the prediction results, we also store the source node ids and destination node ids of the corresponding target edges. So users can map a prediction result back to the right edge according to its source node id and its destination node id.
Currently, the node task and edge task inference scripts can only save prediction results for one node or edge type. This PR makes it more general that the scripts can save prediction results from multiple node types and edge types. First PR of awslabs#508 By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice. --------- Co-authored-by: Xiang Song <[email protected]>
Change the way how node embeddings are saved
From
PATH_TO_EMB:
|- emb_info.json
|- ntype0_emb.part00000.bin
|- ...
|- ntype1_emb.part00000.bin
|- ...
To
PATH_TO_EMB:
|- emb_info.json
|- ntype0
|- emb.part00000.bin
|- emb.part00001.bin
|- ...
|- ntype1
|- emb.part00000.bin
*Issue #, if available:*
awslabs#508
By submitting this pull request, I confirm that you can use, modify,
copy, and redistribute this contribution, under the terms of your
choice.
---------
Co-authored-by: Xiang Song <[email protected]>
Originally, we save prediction results only. However, unlink node prediction tasks, it is impossible to find the corresponding edges for the edge prediction results. In this PR, other than storing the prediction results, we also store the source node ids and destination node ids of the corresponding target edges. So users can map a prediction result back to the right edge according to its source node id and its destination node id. Second PR of awslabs#508 By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice. --------- Co-authored-by: Xiang Song <[email protected]>
Originally, the saving node prediction result process saves the prediction results of all the nodes of target node type(s) regardless of whether a node belongs to an inference set or not. The same thing happends for the saving node embedding process. In this PR, we only save the prediction results and node embeddings of target nodes in the inference set. Third PR of awslabs#508 Also Fix: awslabs#519 By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice. --------- Co-authored-by: Xiang Song <[email protected]>
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.
Update graphstrom.gconstruct.remap_result.py to remap node prediction results.
Will add support for remapping embeddings.
#508
By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.