[CK TILE GEMM] Refactor block_scale_gemm examples

- Update README
This commit is contained in:
Cong Ma
2025-11-10 12:34:38 -05:00
parent 8818018d0c
commit 9debcc1a55
2 changed files with 7 additions and 6 deletions

View File

@@ -52,9 +52,9 @@ args:
-stride_q Tensor AQ stride (default:0)
-stride_b Tensor B stride (default:0)
-stride_c Tensor C stride (default:0)
-v 0. No validation, 1. Validation on CPU, 2. Validation on GPU (default:1)
-prec Data type. For AQuant: fp8/bf8/i4fp8/i4bf8, For Bquant: fp8/bf8/fp8i4/bf8i4 (default:fp8)
-warmup Number of iterations before benchmark the kernel (default:50)
-v 0: No validation, 1: Validation on CPU, 2: Validation on GPU (default:1)
-prec Data type. For AQuant: fp8, bf8, i4fp8, or i4bf8; for Bquant: fp8, bf8, fp8i4, or bf8i4 (default for both AQuant and Bquant: fp8)
-warmup Number of iterations before benchmarking the kernel (default:50)
-repeat Number of iterations to benchmark the kernel (default:1000)
-timer gpu:gpu timer, cpu:cpu timer (default:gpu)
-split_k SplitK value (default:1)

View File

@@ -28,11 +28,12 @@ auto create_args(int argc, char* argv[])
.insert("stride_q", "0", "Tensor AQ stride")
.insert("stride_b", "0", "Tensor B stride")
.insert("stride_c", "0", "Tensor C stride")
.insert("v", "1", "0. No validation, 1. Validation on CPU, 2. Validation on GPU")
.insert("v", "1", "0: No validation, 1: Validation on CPU, 2: Validation on GPU")
.insert("prec",
"fp8",
"Data type. For AQuant: fp8/bf8/i4fp8/i4bf8, For Bquant: fp8/bf8/fp8i4/bf8i4")
.insert("warmup", "50", "Number of iterations before benchmark the kernel")
"Data type. For AQuant: fp8, bf8, i4fp8, or i4bf8; for Bquant: fp8, bf8, fp8i4, "
"or bf8i4")
.insert("warmup", "50", "Number of iterations before benchmarking the kernel")
.insert("repeat", "1000", "Number of iterations to benchmark the kernel")
.insert("timer", "gpu", "gpu:gpu timer, cpu:cpu timer")
.insert("split_k", "1", "SplitK value")