diff --git a/benchmarks/bench_decode_moe.py b/benchmarks/bench_decode_moe.py index 5662179270..8ee26c37de 100644 --- a/benchmarks/bench_decode_moe.py +++ b/benchmarks/bench_decode_moe.py @@ -39,6 +39,7 @@ Run (all three backends, one server per backend): ... --model /path/to/model --backend offload,cpu,hybrid --json out.json + halo """ from __future__ import annotations