Searched defs:BaseSchedulerNode (Results 1 – 9 of 9) sorted by relevance
/aosp_15_r20/external/pytorch/torch/_inductor/ |
H A D | scheduler.py | 158 class BaseSchedulerNode: class 321 self, name_to_fused_node: Dict[str, BaseSchedulerNode] argument 553 def is_materialized(buf: str, snodes: Sequence[BaseSchedulerNode]) -> bool: 710 def __init__(self, node1: BaseSchedulerNode, node2: BaseSchedulerNode) -> None: 752 node: BaseSchedulerNode, 753 name_to_fused_node: Dict[str, BaseSchedulerNode], argument 1038 def refresh_group_node_dependencies(group_snode: BaseSchedulerNode) -> None: 1055 group_snode: BaseSchedulerNode, 1057 snodes: List[BaseSchedulerNode], 1087 cls, node1: BaseSchedulerNode, node2: BaseSchedulerNode [all …]
|
H A D | comms.py | 32 def sink_waits(snodes: List[BaseSchedulerNode]) -> List[BaseSchedulerNode]: 41 def raise_comms(snodes: List[BaseSchedulerNode]) -> List[BaseSchedulerNode]: 51 snodes: List[BaseSchedulerNode], 70 snodes: List[BaseSchedulerNode], 224 nodes: List[BaseSchedulerNode], name_to_buf, name_to_fused_node 255 def estimate_op_runtime(snode: BaseSchedulerNode) -> float: 318 snodes: List[BaseSchedulerNode], 468 snodes: List[torch._inductor.scheduler.BaseSchedulerNode],
|
H A D | debug.py | 54 nodes: List[BaseSchedulerNode], 99 def create_fx_from_snodes(snodes: List[BaseSchedulerNode]) -> fx.Graph: 150 def in_output(snode: Union[BaseSchedulerNode, FusedSchedulerNode]) -> bool:
|
/aosp_15_r20/external/pytorch/torch/_inductor/codegen/ |
H A D | triton_combo_kernel.py | 50 nodes: List[BaseSchedulerNode], 52 kernel_map: Dict[BaseSchedulerNode, TritonKernel], 53 node_info_map: Dict[BaseSchedulerNode, Tuple[Any, Any, Any, Any]], 140 List[BaseSchedulerNode], 142 Dict[BaseSchedulerNode, TritonKernel], 145 List[List[BaseSchedulerNode]], 178 node_info: BaseSchedulerNode, 190 subkernel_nodes: List[BaseSchedulerNode], 192 node_info_map: Dict[BaseSchedulerNode, Tuple[Any, Any, Any, Any]], 245 nodes: List[BaseSchedulerNode], [all …]
|
H A D | cuda_combined_scheduling.py | 36 def choose_node_backend(self, node: BaseSchedulerNode) -> BaseScheduling: 43 def can_fuse_vertical(self, node1: BaseSchedulerNode, node2: BaseSchedulerNode): 48 def can_fuse_horizontal(self, node1: BaseSchedulerNode, node2: BaseSchedulerNode): 61 template_node: BaseSchedulerNode, 62 epilogue_nodes: Sequence[BaseSchedulerNode],
|
H A D | cpp.py | 329 cls, node1: BaseSchedulerNode, node2: BaseSchedulerNode, outer_loop_fusion_depth 4289 def get_call_ranges(node: BaseSchedulerNode): 4476 def is_cpp_template(self, node: BaseSchedulerNode) -> bool: 4483 template_node: BaseSchedulerNode, 4484 epilogue_nodes: Sequence[BaseSchedulerNode],
|
H A D | simd.py | 1514 subkernel_nodes: List[BaseSchedulerNode],
|
/aosp_15_r20/external/pytorch/torch/_inductor/codegen/rocm/ |
H A D | rocm_cpp_scheduling.py | 34 def is_rocm_cpp_template(node: BaseSchedulerNode) -> bool: 40 self, node1: BaseSchedulerNode, node2: BaseSchedulerNode 76 template_node: BaseSchedulerNode, 77 epilogue_nodes: Sequence[BaseSchedulerNode],
|
/aosp_15_r20/external/pytorch/torch/_inductor/codegen/cuda/ |
H A D | cuda_cpp_scheduling.py | 39 def is_cuda_cpp_template(node: BaseSchedulerNode) -> bool: 45 self, node1: BaseSchedulerNode, node2: BaseSchedulerNode 81 template_node: BaseSchedulerNode, 82 epilogue_nodes: Sequence[BaseSchedulerNode],
|