[MoE Refactor][17/N] Apply Refactor to Bf16 (#31827)

Signed-off-by: Yongye Zhu <zyy1102000@gmail.com>
Co-authored-by: Robert Shaw <114415538+robertgshaw2-redhat@users.noreply.github.com>
This commit is contained in:
Yongye Zhu
2026-01-15 12:53:40 -08:00
committed by GitHub
co-authored by Robert Shaw
parent 8c11001ba2
commit 31c29257c8
12 changed files with 257 additions and 87 deletions
@@ -0,0 +1,5 @@
model_name: "Qwen/Qwen3-30B-A3B"
accuracy_threshold: 0.88
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --data-parallel-size 2 --enable-expert-parallel"
@@ -8,3 +8,4 @@ Qwen3-30B-A3B-Fp8-AutoFp8-deepgemm.yaml
Qwen3-30B-A3B-Fp8-CT-Block-deepgemm-deepep-ht.yaml
Qwen3-30B-A3B-Fp8-CT-Block-deepgemm-deepep-ll.yaml
Qwen3-30B-A3B-Fp8-CT-Block-deepgemm.yaml
Qwen3-30B-A3B-BF16-triton.yaml
@@ -0,0 +1,7 @@
model_name: "meta-llama/Llama-4-Scout-17B-16E-Instruct"
accuracy_threshold: 0.92
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2 --enable-expert-parallel"
env:
VLLM_USE_FLASHINFER_MOE_FP16: "1"
@@ -0,0 +1,6 @@
model_name: "meta-llama/Llama-4-Scout-17B-16E-Instruct"
accuracy_threshold: 0.92
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2"
@@ -0,0 +1,7 @@
model_name: "mistralai/Mixtral-8x7B-v0.1"
accuracy_threshold: 0.58
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2 --enable-expert-parallel"
env:
VLLM_USE_FLASHINFER_MOE_FP16: "1"
@@ -0,0 +1,5 @@
model_name: "mistralai/Mixtral-8x7B-v0.1"
accuracy_threshold: 0.58
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2"
@@ -0,0 +1,7 @@
model_name: "Qwen/Qwen3-30B-A3B"
accuracy_threshold: 0.88
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2 --enable-expert-parallel"
env:
VLLM_USE_FLASHINFER_MOE_FP16: "1"
@@ -0,0 +1,5 @@
model_name: "Qwen/Qwen3-30B-A3B"
accuracy_threshold: 0.88
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2"
@@ -11,3 +11,7 @@ Qwen3-30B-A3B-NvFp4-ModelOpt-marlin.yaml
Qwen3-30B-A3B-NvFp4-ModelOpt-fi-trtllm.yaml
Qwen3-30B-A3B-NvFp4-ModelOpt-fi-cutlass.yaml
Qwen3-30B-A3B-NvFp4-ModelOpt-fi-cutlass-dp-ep.yaml
Llama-4-Scout-BF16-fi-cutlass.yaml
Llama-4-Scout-BF16-triton.yaml
Mixtral-8x7B-BF16-fi-cutlass.yaml
Mixtral-8x7B-BF16-triton.yaml
@@ -11,3 +11,5 @@ Qwen3-30B-A3B-Fp8-CT-Channel-vllm-cutlass.yaml
Llama-4-Scout-Fp8-ModelOpt-fi-cutlass.yaml
Llama-4-Scout-Fp8-ModelOpt-marlin.yaml
Llama-4-Scout-Fp8-ModelOpt-triton.yaml
Qwen3-30B-A3B-BF16-fi-cutlass.yaml
Qwen3-30B-A3B-BF16-triton.yaml