[LLM:Bugfix] Honor --chunk_size when generating QNN prefill IO#4610
Closed
LudovicoYIN wants to merge 2 commits into
Closed
[LLM:Bugfix] Honor --chunk_size when generating QNN prefill IO#4610LudovicoYIN wants to merge 2 commits into
LudovicoYIN wants to merge 2 commits into
Conversation
Contributor
Author
|
Closing this PR because it unintentionally included unrelated SeqMSE changes. I will reopen a clean PR that only contains the chunk-size fix. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
This PR fixes a bug in
transformers/llm/export/npu/generate_llm_qnn.pywhere the user-provided--chunk_sizewas not honored when generating prefill IO.Before this change,
convert_llm()always called:which meant the prefill QNN graph could still be generated against 128-token IO even when
--chunk_size 256/512was requested.This can make the generated QNN artifact inconsistent with runtime configuration and lead to misleading evaluation results, especially for
ppl_eval.Module
LLM
Type
Checklist
[Module:Type] Descriptionformat