FM / _C

/usr/local/lib64/python3.6/site-packages/torch/_C/

/usr/local/lib64/python3.6/site-packages/torch/_C

Name	Size	Mode	Actions
_autograd.pyi	2744	0644	edit dl rm
_cudnn.pyi	378	0644	edit dl rm
_distributed_autograd.pyi	908	0644	edit dl rm
_distributed_c10d.pyi	9376	0644	edit dl rm
_distributed_rpc.pyi	5630	0644	edit dl rm
_distributed_rpc_testing.pyi	1136	0644	edit dl rm
_functions.pyi	345	0644	edit dl rm
_nn.pyi	2050	0644	edit dl rm
_nvtx.pyi	150	0644	edit dl rm
_onnx.pyi	646	0644	edit dl rm
_VariableFunctions.pyi	141026	0644	edit dl rm
__init__.pyi	103906	0644	edit dl rm

Edit: /usr/local/lib64/python3.6/site-packages/torch/_C/_autograd.pyi (2744B)

from typing import List, Set, Callable
from enum import Enum

# Defined in tools/autograd/init.cpp

class ProfilerState(Enum):
    Disable = ...
    CPU = ...
    CUDA = ...
    NVTX = ...
    KINETO = ...
    KINETO_GPU_FALLBACK = ...

class ProfilerActivity(Enum):
    CPU = ...
    CUDA = ...

class DeviceType(Enum):
    CPU = ...
    CUDA = ...
    MKLDNN = ...
    OPENGL = ...
    OPENCL = ...
    IDEEP = ...
    HIP = ...
    FPGA = ...
    ORT = ...
    XLA = ...
    MLC = ...
    HPU = ...
    Meta = ...
    Vulkan = ...
    Metal = ...
    ...

class ProfilerConfig:
    def __init__(
        self,
        state: ProfilerState,
        report_input_shapes: bool,
        profile_memory: bool,
        with_stack: bool,
        with_flops: bool,
        with_modules: bool
    ) -> None: ...
    ...

class ProfilerEvent:
    def cpu_elapsed_us(self, other: ProfilerEvent) -> float: ...
    def cpu_memory_usage(self) -> int: ...
    def cuda_elapsed_us(self, other: ProfilerEvent) -> float: ...
    def cuda_memory_usage(self) -> int: ...
    def device(self) -> int: ...
    def handle(self) -> int: ...
    def has_cuda(self) -> bool: ...
    def is_remote(self) -> bool: ...
    def kind(self) -> int: ...
    def name(self) -> str: ...
    def node_id(self) -> int: ...
    def sequence_nr(self) -> int: ...
    def shapes(self) -> List[List[int]]: ...
    def thread_id(self) -> int: ...
    def flops(self) -> float: ...
    def is_async(self) -> bool: ...
    ...

class _KinetoEvent:
    def name(self) -> str: ...
    def device_index(self) -> int: ...
    def start_us(self) -> int: ...
    def duration_us(self) -> int: ...
    def is_async(self) -> bool: ...
    ...

class _ProfilerResult:
    def events(self) -> List[_KinetoEvent]: ...
    def legacy_events(self) -> List[List[ProfilerEvent]]: ...
    def save(self, path: str) -> None: ...

class SavedTensor:
    ...

def _enable_profiler(config: ProfilerConfig, activities: Set[ProfilerActivity]) -> None: ...
def _prepare_profiler(config: ProfilerConfig, activities: Set[ProfilerActivity]) -> None: ...
def _disable_profiler() -> _ProfilerResult: ...
def _profiler_enabled() -> bool: ...
def _add_metadata_json(key: str, value: str) -> None: ...
def kineto_available() -> bool: ...
def _supported_activities() -> Set[ProfilerActivity]: ...
def _enable_record_function(enable: bool) -> None: ...
def _set_empty_test_observer(is_global: bool, sampling_prob: float) -> None: ...
def _register_saved_tensors_default_hooks(pack_hook: Callable, unpack_hook: Callable) -> None: ...
def _reset_saved_tensors_default_hooks() -> None: ...

def _enable_profiler_legacy(config: ProfilerConfig) -> None: ...
def _disable_profiler_legacy() -> List[List[ProfilerEvent]]: ...

from typing import List, Set, Callable from enum import Enum # Defined in tools/autograd/init.cpp class ProfilerState(Enum): Disable = ... CPU = ... CUDA = ... NVTX = ... KINETO = ... KINETO_GPU_FALLBACK = ... class ProfilerActivity(Enum): CPU = ... CUDA = ... class DeviceType(Enum): CPU = ... CUDA = ... MKLDNN = ... OPENGL = ... OPENCL = ... IDEEP = ... HIP = ... FPGA = ... ORT = ... XLA = ... MLC = ... HPU = ... Meta = ... Vulkan = ... Metal = ... ... class ProfilerConfig: def __init__( self, state: ProfilerState, report_input_shapes: bool, profile_memory: bool, with_stack: bool, with_flops: bool, with_modules: bool ) -> None: ... ... class ProfilerEvent: def cpu_elapsed_us(self, other: ProfilerEvent) -> float: ... def cpu_memory_usage(self) -> int: ... def cuda_elapsed_us(self, other: ProfilerEvent) -> float: ... def cuda_memory_usage(self) -> int: ... def device(self) -> int: ... def handle(self) -> int: ... def has_cuda(self) -> bool: ... def is_remote(self) -> bool: ... def kind(self) -> int: ... def name(self) -> str: ... def node_id(self) -> int: ... def sequence_nr(self) -> int: ... def shapes(self) -> List[List[int]]: ... def thread_id(self) -> int: ... def flops(self) -> float: ... def is_async(self) -> bool: ... ... class _KinetoEvent: def name(self) -> str: ... def device_index(self) -> int: ... def start_us(self) -> int: ... def duration_us(self) -> int: ... def is_async(self) -> bool: ... ... class _ProfilerResult: def events(self) -> List[_KinetoEvent]: ... def legacy_events(self) -> List[List[ProfilerEvent]]: ... def save(self, path: str) -> None: ... class SavedTensor: ... def _enable_profiler(config: ProfilerConfig, activities: Set[ProfilerActivity]) -> None: ... def _prepare_profiler(config: ProfilerConfig, activities: Set[ProfilerActivity]) -> None: ... def _disable_profiler() -> _ProfilerResult: ... def _profiler_enabled() -> bool: ... def _add_metadata_json(key: str, value: str) -> None: ... def kineto_available() -> bool: ... def _supported_activities() -> Set[ProfilerActivity]: ... def _enable_record_function(enable: bool) -> None: ... def _set_empty_test_observer(is_global: bool, sampling_prob: float) -> None: ... def _register_saved_tensors_default_hooks(pack_hook: Callable, unpack_hook: Callable) -> None: ... def _reset_saved_tensors_default_hooks() -> None: ... def _enable_profiler_legacy(config: ProfilerConfig) -> None: ... def _disable_profiler_legacy() -> List[List[ProfilerEvent]]: ...