[CI] Tiny speed up VLM CI (#14517)

Co-authored-by: Brayden Zhong <b8zhong@users.noreply.github.com>
2025-12-07 13:30:41 -08:00
parent f6423b626c
commit 3b47973af8
3 changed files with 5 additions and 5 deletions
--- a/python/sglang/test/mmmu_vlm_mixin.py
+++ b/python/sglang/test/mmmu_vlm_mixin.py
@@ -53,7 +53,7 @@ class MMMUVLMMixin(ABC):
        model = "openai_compatible"
        tp = 1
        tasks = "mmmu_val"
-        batch_size = 32
+        batch_size = 64
        log_suffix = "openai_compatible"
        os.makedirs(output_path, exist_ok=True)

@@ -137,7 +137,7 @@ class MMMUVLMMixin(ABC):
                other_args=[
                    "--trust-remote-code",
                    "--cuda-graph-max-bs",
-                    "32",
+                    "64",
                    "--enable-multimodal",
                    "--mem-fraction-static",
                    str(self.parsed_args.mem_fraction_static),  # Use class variable