Fix lint issue and break the long line.

Author: hongtaozhang

superbench/benchmarks/micro_benchmarks/cuda_gemm_flops_performance.py

@@ -24,7 +24,8 @@ def __init__(self, name, parameters=''):
 
         self._bin_name = 'cutlass_profiler'
 
-        # TODO - To support more architecutres, currently only support compute capability = 7.0, 7.5, 8.0, 8.6, 8.9, 9.0, 10.0
+        # TODO - To support more architectures,
+        # currently only support compute capability = 7.0, 7.5, 8.0, 8.6, 8.9, 9.0, 10.0
         self.__kernel_map = {
             7.0: {
                 'fp64': 'cutlass_simt_dgemm_128x128_8x2_*',