update tuned factors for fp16

Author: masahi

tests/python/unittest/test_tir_schedule_tensorize_ldmatrix_mma.py

@@ -288,7 +288,7 @@ def index_map(i, j):
     k_inner = 16
     in_dtype = "float16"
     out_dtype = "float16"
-    i_factors, j_factors, k_factors = [4, 8, 2, 4, 1], [1, 64, 2, 1, 2], [128, 2, 1]
+    i_factors, j_factors, k_factors = [16, 2, 1, 4, 2], [16, 2, 2, 1, 4], [128, 2, 1]
 
     timer = run_test(
         k_inner,