MCPcopy Create free account
hub / github.com/NVIDIA/TensorRT-LLM / pad_up

Function pad_up

tensorrt_llm/math_utils.py:19–20  ·  view source on GitHub ↗
(x: int, y: int)

Source from the content-addressed store, hash-verified

17
18
19def pad_up(x: int, y: int) -> int:
20 return ((x + y - 1) // y) * y

Callers 8

test_fp4_row_linear_allreduceFunction · 0.90
test_fp4_linearFunction · 0.90
test_fp4_linear_cute_dslFunction · 0.90
fp4_linear_perf_testFunction · 0.90
compute_swizzled_sf_shapeFunction · 0.90
_Function · 0.90
calculate_required_workspace_sizeMethod · 0.90
calculate_required_workspace_sizeMethod · 0.90

Calls

no outgoing calls

Tested by 4

test_fp4_row_linear_allreduceFunction · 0.72
test_fp4_linearFunction · 0.72
test_fp4_linear_cute_dslFunction · 0.72
fp4_linear_perf_testFunction · 0.72