Add encoding propagation as an experimental feature #2945
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
Signed-off-by: Kyunggeun Lee <quic_kyunggeu@quicinc.com>
Signed-off-by: Kyunggeun Lee <quic_kyunggeu@quicinc.com>
Signed-off-by: Kyunggeun Lee <quic_kyunggeu@quicinc.com>
Signed-off-by: Kyunggeun Lee <quic_kyunggeu@quicinc.com>
quic-kyunggeu
force-pushed
the
encoding_propagation
branch
from
a9bbc42
to
5862c2a
Compare
quic-akhobare
approved these changes
May 24, 2024
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Outdated suggestions cannot be applied.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.
Below is how encoding unification works in plain english
Given a
module, for each input of themodule, find the "effective producer" of the input.An "effective producer" of a quantized tensor is the layer which holds the output encoding associated with the quantized tensor
Case 1. If the "effective producer" exists:
Set effective_producer.output_quantizers[i] := module.output_quantizers[0] where i is the index of the output quantizer associated with the input
Case 2. If "effective producer" doesn't exist:
This means the input has been never quantized before. Find the consumer of the input and set
consumer.input_quantizers[i] := module.output_quantizers[0] where i is the index of the input quantizer associated with the input