[Core] Optimize request output tokens putting back implementation to reduce overhead #56
bench_test.yml
on: pull_request
cancel_previous_workflows
7s
bench_tests
10m 28s