Remove VLLM_TEST_FORCE_FP8_MARLIN to replace with linear_backend/moe_backend (#52182)

Signed-off-by: mgoin <mgoin64@gmail.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
This commit is contained in:
Misha Goin
2026-08-18 12:58:04 -04:00
committed by GitHub
co-authored by OpenAI Codex
parent 88b2bff2c6
commit bca7bea240
15 changed files with 46 additions and 66 deletions
@@ -2,6 +2,4 @@ model_name: "nvidia/Llama-4-Scout-17B-16E-Instruct-FP8"
accuracy_threshold: 0.92
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2"
env:
VLLM_TEST_FORCE_FP8_MARLIN: "1"
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2 --linear-backend marlin --moe-backend marlin"
@@ -2,6 +2,4 @@ model_name: "Qwen/Qwen3-Coder-30B-A3B-Instruct-FP8"
accuracy_threshold: 0.88
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2"
env:
VLLM_TEST_FORCE_FP8_MARLIN: "1"
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2 --linear-backend marlin --moe-backend marlin"
@@ -2,6 +2,4 @@ model_name: "RedHatAI/Qwen3-30B-A3B-FP8-block"
accuracy_threshold: 0.85
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2"
env:
VLLM_TEST_FORCE_FP8_MARLIN: "1"
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2 --linear-backend marlin --moe-backend marlin"
@@ -2,6 +2,4 @@ model_name: "RedHatAI/Qwen3-30B-A3B-FP8-dynamic"
accuracy_threshold: 0.85
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2"
env:
VLLM_TEST_FORCE_FP8_MARLIN: "1"
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2 --linear-backend marlin --moe-backend marlin"
@@ -2,6 +2,4 @@ model_name: "RedHatAI/Qwen3-30B-A3B-NVFP4"
accuracy_threshold: 0.88
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2"
env:
VLLM_TEST_FORCE_FP8_MARLIN: "1"
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2 --linear-backend marlin --moe-backend marlin"
@@ -2,6 +2,4 @@ model_name: "nvidia/Qwen3-30B-A3B-NVFP4"
accuracy_threshold: 0.88
num_questions: 1319
num_fewshot: 5
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2"
env:
VLLM_TEST_FORCE_FP8_MARLIN: "1"
server_args: "--enforce-eager --max-model-len 8192 --tensor-parallel-size 2 --linear-backend marlin --moe-backend marlin"