Replace sglang.srt.layers.quantization.scalar_types with sgl_kernel.scalar_type (#8951)

2025-08-14 10:41:41 +08:00
parent 6b7c24712c
commit a669bc2f74
8 changed files with 44 additions and 362 deletions
--- a/python/sglang/srt/layers/quantization/compressed_tensors/compressed_tensors_moe.py
+++ b/python/sglang/srt/layers/quantization/compressed_tensors/compressed_tensors_moe.py
@@ -16,7 +16,6 @@ from sglang.srt.layers.quantization.fp8_kernel import is_fp8_fnuz, scaled_fp8_qu
 from sglang.srt.layers.quantization.fp8_utils import normalize_e4m3fn_to_e4m3fnuz
 from sglang.srt.layers.quantization.utils import (
    all_close_1d,
-    cpu_has_amx_support,
    per_tensor_dequantize,
    replace_parameter,
 )