forked from vllm-project/vllm
-
Notifications
You must be signed in to change notification settings - Fork 97
Pull requests: HabanaAI/vllm-fork
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[deepseek_r1] add scripts for benchmark throughput and serving
#1288
opened May 21, 2025 by
yangulei
Loading…
Increase the default value of VLLM_MOE_SLICE_LENGTH to 100k
#1287
opened May 21, 2025 by
czhu15
Loading…
[SW-225565] Enable traingular softmax with merged prefill
#1278
opened May 20, 2025 by
kamil-kaczor
•
Draft
fix the incorrect output_tokens for penalty calculation in the sampler when delayed sampling is enabled
#1199
opened May 6, 2025 by
ccrhx4
Loading…
Previous Next
ProTip!
Exclude everything labeled
bug
with -label:bug.