A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 11, 17:30 UTC
[CI/Test] Fix FP8 per-tensor quant test reference scale shape (#30352)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Lucas Wilkinson committed
Dec 9, 2025 at 13:52 UTC
b37bf51e7594133772f4f06446dd5aa08aa5be0e
1 file changed
+1
-1
tests/kernels/quant_utils.py
+1
-1
@@ -103,7 +103,7 @@ def ref_dynamic_per_tensor_fp8_quant(
103
.clamp(fp8_traits_min, fp8_traits_max)
104
.to(FP8_DTYPE)
105
)
106
- return ref_out, ref_scale.view((1, 1))
106
+ return ref_out, ref_scale.view(1)
107
108
109
def native_w8a8_block_matmul(