修复block-wise 用triton报错的bug

359c16db · SAC_fanth · 9d44744c · 359c16db
Commit 359c16db authored Feb 25, 2026 by SAC_fanth
Show whitespace changes
Inline Side-by-side

Showing with 1 addition and 0 deletions

vllm/model_executor/layers/quantization/utils/fp8_utils.py vllm/model_executor/layers/quantization/utils/fp8_utils.py +1 -0

No files found.
--- a/vllm/model_executor/layers/quantization/utils/fp8_utils.py
+++ b/vllm/model_executor/layers/quantization/utils/fp8_utils.py
@@ -526,6 +526,7 @@ class W8A8BlockFp8LinearOp:

    def _run_triton(
        self,
+        out_features: int,
        input_2d: torch.Tensor,
        weight: torch.Tensor,
        weight_scale: torch.Tensor,