Honor MAGPIE_EVAL_TOKENIZED_REQUESTS in remote-compat lm_eval model_args - #74
Open
weilei0120 wants to merge 1 commit into
Open
Honor MAGPIE_EVAL_TOKENIZED_REQUESTS in remote-compat lm_eval model_args#74weilei0120 wants to merge 1 commit into
weilei0120 wants to merge 1 commit into
Conversation
lm_eval defaults to token-id-array prompts, which a direct sglang server accepts but a PD-disaggregated sglang_router rejects with HTTP 422, collapsing the accuracy eval. Let the model_args line append tokenized_requests when MAGPIE_EVAL_TOKENIZED_REQUESTS is set (e.g. =false on PD to send string prompts); unset/empty is byte-for-byte the previous behaviour. Co-authored-by: Cursor <cursoragent@cursor.com>
Collaborator
|
Thank you! @weilei0120 Could you also add this to doc? |
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
lm_eval defaults to token-id-array prompts, which a direct sglang server accepts but a PD-disaggregated sglang_router rejects with HTTP 422, collapsing the accuracy eval. Let the model_args line append tokenized_requests when MAGPIE_EVAL_TOKENIZED_REQUESTS is set (e.g. =false on PD to send string prompts); unset/empty is byte-for-byte the previous behaviour.