[CI] Fix LM Eval Large Models (H100) (#32423)

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>

[CI] Fix LM Eval Large Models (H100) (#32423)
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
bcf2333c · Matthew Bonanni · GitHub · 83239ff1 · bcf2333c
Unverified Commit bcf2333c authored Jan 15, 2026 by Matthew Bonanni Committed by GitHub Jan 16, 2026
Show whitespace changes
Inline Side-by-side

Showing with 3 additions and 1 deletion

vllm/model_executor/layers/quantization/input_quant_fp8.py vllm/model_executor/layers/quantization/input_quant_fp8.py +3 -1

No files found.
--- a/vllm/model_executor/layers/quantization/input_quant_fp8.py
+++ b/vllm/model_executor/layers/quantization/input_quant_fp8.py
@@ -98,7 +98,9 @@ class QuantFP8(CustomOp):
            num_token_padding=self.num_token_padding,
            scale_ub=scale_ub,
            use_per_token_if_dynamic=self.use_per_token_if_dynamic,
-            group_shape=self.group_shape if self.static else None,
+            group_shape=(self.group_shape.row, self.group_shape.col)
+            if self.static
+            else None,
        )

    def forward_hip(