use srt VocabParallelEmbedding (#3252) #5148
pr-test.yml
on: push
Matrix: unit-test-backend-1-gpu
unit-test-frontend
2m 20s
unit-test-backend-2-gpu
10m 16s
performance-test-1-gpu-part-1
11m 32s
performance-test-1-gpu-part-2
11m 35s
performance-test-2-gpu
10m 39s
accuracy-test-1-gpu
5m 31s
accuracy-test-2-gpu
6m 33s
finish
0s