MCPcopy Create free account

hub / github.com/bytedance/flux / functions

Functions784 in github.com/bytedance/flux

⨍ Functions 784◇ Types & classes 309
↓ 2 callersFunction_rand
(shape, dtype)
test/test_gemm_rs.py:312
↓ 2 callersFunction_run
(local_m)
test/test_ag_kernel_functional.py:93
↓ 2 callersMethodas_python_args
(self)
pynvshmem/tools/autogen_pybind.py:177
↓ 2 callersFunctionatomic_add_release_dev
include/flux/cuda/cuda_common_device.hpp:155
↓ 2 callersFunctionbsr_reduce
src/reduce_scatter/ths_op/helper_ops.cc:37
↓ 2 callersFunctioncheck_nvshmem_init
src/ths_op/flux_shm.cc:42
↓ 2 callersMethodcopy_local
(self, input: torch.Tensor)
python/flux/ag_kernel_crossnode.py:176
↓ 2 callersFunctioncreate_nccl_comm_with_processgroup
src/ths_op/topo_utils.cc:262
↓ 2 callersMethoddefault_evt_d
include/flux/cuda/gemm_impls/gemm_v2_impl.hpp:176
↓ 2 callersMethoddefault_gemm_kernel
include/flux/cuda/gemm_impls/gemm_v2_impl.hpp:250
↓ 2 callersMethoddefault_get_callback_args
include/flux/cuda/gemm_impls/gemm_v2_impl.hpp:388
↓ 2 callersMethoddefault_get_epilogue_args
include/flux/cuda/gemm_impls/gemm_v3_impl.hpp:77
↓ 2 callersFunctionenum_to_string
////////////////////////////////////////////////////////// Print //////////////////////////////////////////////////////////
include/flux/flux.h:676
↓ 2 callersMethodevt_d
include/flux/cuda/gemm_impls/gemm_v2_impl.hpp:165
↓ 2 callersFunctionfn
()
test/test_gemm_only.py:81
↓ 2 callersMethodforward
src/all_gather/ths_op/all_gather_gemm_kernel.cc:639
↓ 2 callersMethodforward_impl
src/comm_none/ths_op/gemm_only.cc:127
↓ 2 callersMethodforward_impl
src/all_gather/ths_op/all_gather_gemm_kernel.cc:314
↓ 2 callersMethodforward_impl
src/reduce_scatter/ths_op/gemm_reduce_scatter.cc:571
↓ 2 callersMethodget_args_workspace_size
for device allocation required by Gemm::Arguments
include/flux/gemm_operator_base.h:92
↓ 2 callersMethodget_current_work
src/all_gather/sm90_all_gather_gemm_tile_scheduler.hpp:136
↓ 2 callersMethodget_current_work
src/reduce_scatter/tile_scheduler/sm90_tile_scheduler_reduce_scatter.hpp:94
↓ 2 callersFunctionget_gpu_bus_id
src/ths_op/topo_utils.cc:87
↓ 2 callersFunctionget_intra_node_pg_group
(tp_group: torch.distributed.ProcessGroup, nnodes: int)
python/flux/gemm_rs_sm80.py:30
↓ 2 callersMethodget_latest_prof_result
src/ths_op/ths_op.cc:169
↓ 2 callersFunctionget_processs_group_devices
src/ths_op/topo_utils.cc:236
↓ 2 callersFunctionget_public_version
()
setup.py:67
↓ 2 callersFunctionget_ring_mode
src/all_gather/ths_op/all_gather_types.h:20
↓ 2 callersMethodget_runtime_gemm_hparams
include/flux/gemm_operator_base.h:114
↓ 2 callersMethodget_stride_c
include/flux/cuda/gemm_impls/gemm_v2_impl.hpp:375
↓ 2 callersFunctionglobal_timer
include/flux/cuda/cuda_common_device.hpp:32
↓ 2 callersFunctionhas_nvlink
src/ths_op/topo_utils.cc:340
↓ 2 callersFunctioninitialize_topo
src/ths_op/topo_utils.cc:300
↓ 2 callersMethodis_producer_load_needed
src/reduce_scatter/sm90_epilogue_evt.hpp:173
↓ 2 callersFunctionis_topo_initialized
src/ths_op/topo_utils.cc:294
↓ 2 callersMethodlazy_init_barrier_buffer
src/reduce_scatter/ths_op/gemm_reduce_scatter.cc:140
↓ 2 callersMethodlazy_init_cpp_op
(self, m_dim: int)
python/flux/gemm_rs_sm80.py:118
↓ 2 callersMethodlazy_init_cpp_op
(self, full_m: int)
python/flux/ag_gemm.py:112
↓ 2 callersMethodlazy_init_cpp_op
(self, full_m: int)
python/flux/ag_kernel_crossnode.py:139
↓ 2 callersMethodlazy_init_gemm_buffer
src/comm_none/ths_op/gemm_only.cc:226
↓ 2 callersMethodlazy_init_gemm_buffer
src/all_gather/ths_op/all_gather_gemm_kernel.cc:225
↓ 2 callersMethodlazy_init_gemm_buffer
src/all_gather/ths_op/all_gather_gemm_kernel_crossnode.cc:281
↓ 2 callersMethodlazy_init_gemm_buffer
src/reduce_scatter/ths_op/gemm_reduce_scatter.cc:187
↓ 2 callersFunctionperf_gemm
(iters: int, name: str, fn: callable)
test/test_gemm_only.py:50
↓ 2 callersMethodreset_signals
(self)
python/flux/ag_kernel_crossnode.py:173
↓ 2 callersFunctionrun_nccl
src/all_gather/ths_op/all_gather_gemm_kernel_crossnode.cc:92
↓ 2 callersMethodtb_swizzle
src/all_gather/gemm_v2_ag_kernel.hpp:49
↓ 2 callersMethodtb_swizzle
include/flux/cuda/gemm_impls/gemm_v2_impl.hpp:93
↓ 2 callersFunctionthreadid
include/flux/cuda/cuda_common_device.hpp:54
↓ 2 callersMethodtile_scheduler
src/all_gather/gemm_v3_ag_kernel.hpp:63
↓ 2 callersMethodtile_scheduler
src/reduce_scatter/gemm_v3_reduce_scatter.hpp:72
↓ 2 callersFunctionto_cutlass_archtag
include/flux/cuda/cuda_common.h:113
↓ 2 callersMethodto_gemm_args
include/flux/cuda/gemm_impls/gemm_v3_impl.hpp:71
↓ 2 callersFunctiontuple_append
include/flux/flux.h:192
↓ 2 callersFunctiontuple_unpack_cat
include/flux/flux.h:421
↓ 2 callersFunctionwait_eq_sys
src/reduce_scatter/reduce_scatter_kernel.hpp:146
↓ 1 callersMethodAGKernelCrossNode
src/all_gather/ths_op/all_gather_gemm_kernel_crossnode.cc:172
↓ 1 callersFunctionToRuntimeConfig
src/cuda/op_registry_proto_utils.cc:99
↓ 1 callersFunctionToUnifiedGemmHparams
src/cuda/op_registry_proto_utils.cc:115
↓ 1 callersFunctionToUnifiedGemmMata
src/cuda/op_registry_proto_utils.cc:62
↓ 1 callersFunction_async_barrier
()
test/utils.py:102
↓ 1 callersFunction_auto_ring_mode
()
python/flux/ag_kernel_crossnode.py:71
↓ 1 callersFunction_get_flops
()
python/flux/util.py:47
↓ 1 callersFunction_get_torch_fp8_dtypes
()
python/flux/util.py:23
↓ 1 callersMethod_lazy_init_inter_ag_output_buffer
(self, input: torch.Tensor)
python/flux/ag_gemm.py:191
↓ 1 callersFunction_load_deps
()
python/flux/cpp_mod.py:39
↓ 1 callersFunction_stream_raw
(stream: CUDA_STREAM_TYPE)
python/flux/pynvshmem/_wrapper.py:16
↓ 1 callersFunction_visit
(node, funcs)
pynvshmem/tools/autogen_pybind.py:261
↓ 1 callersFunctionadd_continous_kernel
src/reduce_scatter/reduce_scatter_kernel.hpp:219
↓ 1 callersFunctionallgather_cpu
src/ths_op/topo_utils.cc:66
↓ 1 callersMethodas_cpp_arguments
(self)
pynvshmem/tools/autogen_pybind.py:226
↓ 1 callersMethodas_cpp_params
(self)
pynvshmem/tools/autogen_pybind.py:241
↓ 1 callersMethodas_pyi
(self)
pynvshmem/tools/autogen_pybind.py:186
↓ 1 callersMethodas_python_params
(self)
pynvshmem/tools/autogen_pybind.py:183
↓ 1 callersFunctionatomic_store_dev
include/flux/cuda/cuda_common_device.hpp:115
↓ 1 callersFunctionautogen
(nvshmem_home, output_dir)
pynvshmem/tools/autogen_pybind.py:325
↓ 1 callersFunctionblockid
include/flux/cuda/cuda_common_device.hpp:48
↓ 1 callersMethodcollective_epilogue
src/reduce_scatter/gemm_v3_reduce_scatter.hpp:191
↓ 1 callersMethodcopy_unpack_
include/flux/cuda/memory_utils.hpp:402
↓ 1 callersFunctioncreate_tensor
pynvshmem/src/functions.cpp:50
↓ 1 callersFunctioncuda_deps
()
setup.py:135
↓ 1 callersMethodcuda_stream_wrapper
(self)
pynvshmem/tools/autogen_pybind.py:196
↓ 1 callersFunctioncuda_version
CUDA Toolkit version as a (major, minor) by nvcc --version
setup.py:27
↓ 1 callersMethodcustom_evt_d
src/reduce_scatter/gemm_v2_reduce_scatter.hpp:285
↓ 1 callersMethodcustom_gemm_device
src/reduce_scatter/gemm_v2_reduce_scatter.hpp:276
↓ 1 callersFunctioncutlass_deps
()
setup.py:89
↓ 1 callersMethoddefault_gemm_device
include/flux/cuda/gemm_impls/gemm_v2_impl.hpp:336
↓ 1 callersMethodevt_store_d
src/reduce_scatter/gemm_v2_reduce_scatter.hpp:193
↓ 1 callersMethodfetch
src/reduce_scatter/sm90_reduce_scatter_utils.hpp:227
↓ 1 callersMethodfetch_tail
src/reduce_scatter/sm90_reduce_scatter_utils.hpp:298
↓ 1 callersFunctionfilter_arch
include/flux/gemm_hparams.h:396
↓ 1 callersFunctionfilter_funcs
(funcs)
pynvshmem/tools/autogen_pybind.py:275
↓ 1 callersFunctionfilter_smem
include/flux/gemm_hparams.h:412
↓ 1 callersFunctionflux_cuda_deps
()
setup.py:127
↓ 1 callersMethodforward
(self, input: torch.Tensor, weight: torch.Tensor)
python/flux/ag_kernel_crossnode.py:167
↓ 1 callersMethodgemm_kernel
src/all_gather/gemm_v3_ag_kernel.hpp:72
↓ 1 callersMethodgemm_kernel
src/reduce_scatter/gemm_v3_reduce_scatter.hpp:202
↓ 1 callersMethodgemm_only
( self, input: torch.Tensor, full_input: torch.Tensor, weight: torch.Tensor )
python/flux/ag_kernel_crossnode.py:179
↓ 1 callersMethodgemm_only
src/all_gather/ths_op/all_gather_gemm_kernel.cc:614
↓ 1 callersFunctiongen_tuning_space
()
tools/tune_ag_gemm_kernel.py:67
← previousnext →101–200 of 784, ranked by callers