File size: 3,238 Bytes
9c6594c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 |
import random
import re
import sys
from dataclasses import dataclass, field
from typing import (
Any,
Callable,
Dict,
Iterable,
Iterator,
List,
Optional,
TypeVar,
Union,
)
from .pytorch import IterableDataset
T = TypeVar('T')
def envlookup(m: re.Match) -> str: ...
def envsubst(s: str) -> str: ...
def split_by_node(src: Iterable[T], group: Any = None) -> Iterator[T]: ...
def single_node_only(src: Iterable[T], group: Any = None) -> Iterator[T]: ...
def split_by_worker(src: Iterable[T]) -> Iterator[T]: ...
def expand_urls(urls: str) -> List[str]: ...
def expand_source(source: Union[str, List[str], Iterable], max_urls: int = int(1e9)) -> List[str]: ...
class SimpleShardList(IterableDataset):
urls: List[str]
seed: Optional[Union[int, bool]]
def __init__(self, urls: Union[str, List[str]], seed: Optional[Union[int, bool]] = None) -> None: ...
def __len__(self) -> int: ...
def __iter__(self) -> Iterator[Dict[str, str]]: ...
def resampled_(src: Iterable[T], n: int = sys.maxsize) -> Iterator[T]: ...
resampled: Any # pipelinefilter return type
def non_empty(src: Iterable[T]) -> Iterator[T]: ...
@dataclass
class MSSource:
name: str = ""
perepoch: int = -1
resample: bool = False
urls: List[str] = field(default_factory=list)
default_rng: random.Random
def expand(s: str) -> str: ...
class ResampledShards(IterableDataset):
urls: List[str]
nshards: int
worker_seed: Callable
deterministic: bool
seed: int
epoch: int
rng: random.Random
def __init__(
self,
urls: Union[str, List[str], Iterable],
nshards: int = sys.maxsize,
seed: int = 0,
worker_seed: Optional[Callable] = None,
deterministic: bool = False,
max_urls: int = int(1e6),
empty_check: bool = True,
) -> None: ...
def __iter__(self) -> Iterator[Dict[str, str]]: ...
ResampledShardList = ResampledShards
def check_pid_is_running(pid: int) -> bool: ...
def without_last_extension(fname: str) -> str: ...
def get_pid_from_filename(fname: str) -> Optional[int]: ...
class DirectoryShardList(IterableDataset):
path: str
poll: int
pattern: str
mode: str
select: str
fate: Any
timeout: float
def __init__(
self,
path: str,
pattern: str = "*.{tar,tgz,tar.tgz}",
poll: int = 1,
timeout: float = 1e12,
mode: str = "resample",
select: str = "random",
fate: Any = None,
) -> None: ...
def recycle(self, activename: str) -> None: ...
def cleanup_files_without_processes(self) -> None: ...
def __iter__(self) -> Iterator[Dict[str, str]]: ...
class MultiShardSample(IterableDataset):
epoch: int
rng: random.Random
sources: List[MSSource]
# Removed obsolete decorator to avoid type errors
def __init__(self, fname: Union[str, Dict[str, Any]]) -> None: ...
def parse_spec(self, fname: Union[str, Dict[str, Any]]) -> None: ...
def set_epoch(self, seed: int) -> None: ...
def get_shards_for_epoch(self) -> List[str]: ...
def __iter__(self) -> Iterator[Dict[str, str]]: ...
def shardspec(spec: str) -> Union[MultiShardSample, SimpleShardList]: ...
|