Improve XR-Transformer memory efficiency #128
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
jiong-zhang
force-pushed
the
xr-transformer-realtime-tokenizer
branch
from
1ab9a78 to
d6a9d4f
Compare
jiong-zhang
force-pushed
the
xr-transformer-realtime-tokenizer
branch
from
b2d9761 to
f64878e
Compare
jiong-zhang
changed the title
rofuyu
approved these changes
Apr 1, 2022
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Outdated suggestions cannot be applied.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.
Issue #, if available:
Description of changes:
XMCTextDataset, a dataset for XMC problem that tokenizes input text at batch generation time. This avoids pre-tokenizing the whole training corpus which could lead to OOM when number of instances is big.--save-emb-dirargument inpecos.xmc.xtransformer.train. Embeddings should be obtained throughpecos.xmc.xtransformer.encode.--steps-scaleargument inpecos.xmc.xtransformer.train. Setting different training steps for each layer should be done through setting custommax_stepsin theTransformerMatcher.TrainParamsBy submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.