@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / 881a6b011b

Fork 0

A high-throughput and memory-efficient inference and serving engine for LLMs

Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git

Synced Nov 7, 09:00 UTC

[CI] Temporarily Disable Llama4 MoE Refactor Test (#35870)

Signed-off-by: Robert Shaw <114415538+robertgshaw2-redhat@users.noreply.github.com>

Robert Shaw committed Mar 3, 2026 at 13:36 UTC 881a6b011b76bddf159b1a635586064e34e221b0

1 file changed -3

tests/evals/gsm8k/configs/moe-refactor/config-h100.txt

-3

@@ -8,8 +8,5 @@ Qwen3-30B-A3B-Fp8-CT-Block-marlin.yaml

8 Qwen3-30B-A3B-Fp8-CT-Block-triton.yaml

9 Qwen3-30B-A3B-Fp8-CT-Channel-marlin.yaml

10 Qwen3-30B-A3B-Fp8-CT-Channel-vllm-cutlass.yaml

11 -Llama-4-Scout-Fp8-ModelOpt-fi-cutlass.yaml

12 -Llama-4-Scout-Fp8-ModelOpt-marlin.yaml

13 -Llama-4-Scout-Fp8-ModelOpt-triton.yaml

11 Qwen3-30B-A3B-BF16-fi-cutlass.yaml

12 Qwen3-30B-A3B-BF16-triton.yaml