[NPU] solve accuracy problem for stablelm-2-1-6b for npu#17470
Merged
iforgetmyname merged 9 commits intosgl-project:mainfrom Jan 24, 2026
Merged
[NPU] solve accuracy problem for stablelm-2-1-6b for npu#17470iforgetmyname merged 9 commits intosgl-project:mainfrom
iforgetmyname merged 9 commits intosgl-project:mainfrom
Conversation
Contributor
|
Warning You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again! |
Contributor
|
Please run pre-commit and update the PR description |
Contributor
Author
Sure. Thank you. |
Collaborator
|
/tag-and-rerun-ci |
iforgetmyname
approved these changes
Jan 24, 2026
Johnsonms
pushed a commit
to Johnsonms/sglang
that referenced
this pull request
Feb 14, 2026
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
3 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Motivation
Motivation
accuracy fix for stablelm-2-1-6b for npu, previously it was 0.0
Modifications
as follows.
Increase the dtype before performing rope processing to avoid affecting the precision.
Accuracy Tests
20%
##start command
python -m sglang.launch_server
--model-path /xxxxxxxxx/xxxxxxxxx/model path for stablelm-2-1-6b
--trust-remote-code
--mem-fraction-static 0.9
--host 127.0.0.1
--port 8700
--tp-size 1
--device npu
--attention-backend ascend
--max-running-requests 64
--disable-radix-cache
--disable-overlap-schedule
--watchdog-timeout 9000
--base-gpu-id xxxx
--disable-cuda-graph
--enable-torch-compile
--chunked-prefill-size -1 \
##test_gsm8k command
change model path in test/registered/ascend/llm_models/test_ascend_stablelm-2-1_6b.py for stablelm-2-1-6b
python test/registered/ascend/llm_models/test_ascend_stablelm-2-1_6b.py