remove vllm fp8quant from fp8.py (#8937)

aaf0ad8c · Stefan He · GitHub · 361379b5 · aaf0ad8c
Unverified Commit aaf0ad8c authored Aug 07, 2025 by Stefan He Committed by GitHub Aug 07, 2025
Hide whitespace changes
Inline Side-by-side

Showing with 0 additions and 3 deletions

python/sglang/srt/layers/quantization/fp8.py python/sglang/srt/layers/quantization/fp8.py +0 -3

No files found.
--- a/python/sglang/srt/layers/quantization/fp8.py
+++ b/python/sglang/srt/layers/quantization/fp8.py
@@ -98,9 +98,6 @@ if _is_hip and (_use_aiter or _use_hip_int4):
    from aiter.fused_moe import fused_moe
    from aiter.ops.shuffle import shuffle_weight

-if not (_is_cuda or _is_npu or (_is_cpu and _is_cpu_amx_available) or _is_hip):
-    from vllm._custom_ops import scaled_fp8_quant
-

 ACTIVATION_SCHEMES = ["static", "dynamic"]