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 9, 14:45 UTC
[Tests] Remove Duplicates (#33032)
Signed-off-by: Robert Shaw <robshaw@redhat.com> Co-authored-by: Robert Shaw <robshaw@redhat.com>
Robert Shaw committed
Jan 26, 2026 at 00:23 UTC
254db42ede6beb7d3191f50084594c0ce791ce40
3 files changed
-18
tests/evals/gsm8k/configs/moe-refactor/Qwen3-30B-A3B-NvFp4-CT-fi-cutlass-dp-ep.yaml
deleted
-8
@@ -1,8 +0,0 @@
1
-model_name: "RedHatAI/Qwen3-30B-A3B-NVFP4"
2
-accuracy_threshold: 0.88
3
-num_questions: 1319
4
-num_fewshot: 5
5
-server_args: "--enforce-eager --max-model-len 8192 --data-parallel-size 2 --enable-expert-parallel"
6
-env:
7
- VLLM_USE_FLASHINFER_MOE_FP4: "1"
8
- VLLM_FLASHINFER_MOE_BACKEND: "throughput"
tests/evals/gsm8k/configs/moe-refactor/Qwen3-30B-A3B-NvFp4-ModelOpt-fi-cutlass-dp-ep.yaml
deleted
-8
@@ -1,8 +0,0 @@
1
-model_name: "nvidia/Qwen3-30B-A3B-NVFP4"
2
-accuracy_threshold: 0.88
3
-num_questions: 1319
4
-num_fewshot: 5
5
-server_args: "--enforce-eager --max-model-len 8192 --data-parallel-size 2 --enable-expert-parallel"
6
-env:
7
- VLLM_USE_FLASHINFER_MOE_FP4: "1"
8
- VLLM_FLASHINFER_MOE_BACKEND: "throughput"
tests/evals/gsm8k/configs/moe-refactor/config-b200.txt
-2
@@ -5,12 +5,10 @@ Qwen3-30B-A3B-NvFp4-CT-vllm-cutlass.yaml
5
Qwen3-30B-A3B-NvFp4-CT-marlin.yaml
6
Qwen3-30B-A3B-NvFp4-CT-fi-trtllm.yaml
7
Qwen3-30B-A3B-NvFp4-CT-fi-cutlass.yaml
8
-Qwen3-30B-A3B-NvFp4-CT-fi-cutlass-dp-ep.yaml
8
Qwen3-30B-A3B-NvFp4-ModelOpt-vllm-cutlass.yaml
9
Qwen3-30B-A3B-NvFp4-ModelOpt-marlin.yaml
10
Qwen3-30B-A3B-NvFp4-ModelOpt-fi-trtllm.yaml
11
Qwen3-30B-A3B-NvFp4-ModelOpt-fi-cutlass.yaml
13
-Qwen3-30B-A3B-NvFp4-ModelOpt-fi-cutlass-dp-ep.yaml
12
Llama-4-Scout-BF16-fi-cutlass.yaml
13
Llama-4-Scout-BF16-triton.yaml
14
Mixtral-8x7B-BF16-fi-cutlass.yaml