File size: 16,873 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 |
# Copyright The Lightning AI team.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import inspect
from collections.abc import Generator, Iterable, Mapping, Sized
from dataclasses import fields
from typing import Any, Optional, Union
import torch
from lightning_utilities.core.apply_func import is_dataclass_instance
from torch import Tensor
from torch.utils.data import BatchSampler, DataLoader, IterableDataset, RandomSampler, Sampler, SequentialSampler
from typing_extensions import TypeGuard
import pytorch_lightning as pl
from lightning_fabric.utilities.data import (
_reinstantiate_wrapped_cls,
_replace_value_in_saved_args,
has_iterable_dataset,
sized_len,
)
from lightning_fabric.utilities.warnings import PossibleUserWarning
from pytorch_lightning.overrides.distributed import _IndexBatchSamplerWrapper
from pytorch_lightning.trainer.states import RunningStage
from pytorch_lightning.utilities.exceptions import MisconfigurationException
from pytorch_lightning.utilities.rank_zero import WarningCache, rank_zero_warn
BType = Union[Tensor, str, Mapping[Any, "BType"], Iterable["BType"]]
warning_cache = WarningCache()
def _extract_batch_size(batch: BType) -> Generator[Optional[int], None, None]:
if isinstance(batch, Tensor):
if batch.ndim == 0:
yield 1
else:
yield batch.size(0)
elif isinstance(batch, (Iterable, Mapping)) and not isinstance(batch, str):
if isinstance(batch, Mapping):
batch = batch.values()
for sample in batch:
yield from _extract_batch_size(sample)
elif is_dataclass_instance(batch):
for field in fields(batch): # type: ignore[arg-type]
yield from _extract_batch_size(getattr(batch, field.name))
else:
yield None
def extract_batch_size(batch: BType) -> int:
"""Unpack a batch to find a ``torch.Tensor``.
Returns:
``len(tensor)`` when found, or ``1`` when it hits an empty or non iterable.
"""
error_msg = (
"We could not infer the batch_size from the batch. Either simplify its structure"
" or provide the batch_size as `self.log(..., batch_size=batch_size)`."
)
batch_size = None
try:
for bs in _extract_batch_size(batch):
if batch_size is None:
batch_size = bs
elif batch_size != bs:
warning_cache.warn(
"Trying to infer the `batch_size` from an ambiguous collection. The batch size we"
f" found is {batch_size}. To avoid any miscalculations, use `self.log(..., batch_size=batch_size)`."
)
break
except RecursionError:
raise RecursionError(error_msg)
if batch_size is None:
raise MisconfigurationException(error_msg)
return batch_size
def has_len_all_ranks(
dataloader: object,
strategy: "pl.strategies.Strategy",
allow_zero_length_dataloader_with_multiple_devices: bool = False,
) -> TypeGuard[Sized]:
"""Checks if a given object has ``__len__`` method implemented on all ranks."""
local_length = sized_len(dataloader)
if local_length is None:
# __len__ is not defined, skip these checks
return False
total_length = strategy.reduce(torch.tensor(local_length, device=strategy.root_device), reduce_op="sum")
if total_length == 0:
rank_zero_warn(
f"Total length of `{type(dataloader).__name__}` across ranks is zero."
" Please make sure this was your intention."
)
if total_length > 0 and local_length == 0:
dataloader_cls_name = type(dataloader).__name__
if not allow_zero_length_dataloader_with_multiple_devices:
raise RuntimeError(
f"`{dataloader_cls_name}` within local rank has zero length."
" Please make sure that it returns at least 1 batch."
)
rank_zero_warn(
f"Total length of `{dataloader_cls_name}` across ranks is zero, but local rank has zero"
" length. Please be cautious of uneven batch length."
)
if has_iterable_dataset(dataloader):
rank_zero_warn(
"Your `IterableDataset` has `__len__` defined."
" In combination with multi-process data loading (when num_workers > 1),"
" `__len__` could be inaccurate if each worker is not configured independently"
" to avoid having duplicate data."
)
return True
def _update_dataloader(
dataloader: DataLoader, sampler: Union[Sampler, Iterable], mode: Optional[RunningStage] = None
) -> DataLoader:
dl_args, dl_kwargs = _get_dataloader_init_args_and_kwargs(dataloader, sampler, mode)
return _reinstantiate_wrapped_cls(dataloader, *dl_args, **dl_kwargs)
def _get_dataloader_init_args_and_kwargs(
dataloader: DataLoader,
sampler: Union[Sampler, Iterable],
mode: Optional[RunningStage] = None,
) -> tuple[tuple[Any], dict[str, Any]]:
if not isinstance(dataloader, DataLoader):
raise ValueError(f"The dataloader {dataloader} needs to subclass `torch.utils.data.DataLoader`")
was_wrapped = hasattr(dataloader, "__pl_saved_args")
if was_wrapped:
dl_args = dataloader.__pl_saved_args
dl_kwargs = dataloader.__pl_saved_kwargs
arg_names = dataloader.__pl_saved_arg_names
original_dataset = dataloader.__dataset # we have this saved from _wrap_init
else:
# get the dataloader instance attributes
attrs = {k: v for k, v in vars(dataloader).items() if not k.startswith("_")}
# We cannot be 100% sure the class sets dataset argument. Let's set it to None to be safe
# and hope we can get it from the instance attributes
original_dataset = None
# not part of `vars`
attrs["multiprocessing_context"] = dataloader.multiprocessing_context
arg_names = ()
# get the dataloader instance `__init__` parameters
params = dict(inspect.signature(dataloader.__init__).parameters) # type: ignore[misc]
has_variadic_kwargs = any(p.kind is p.VAR_KEYWORD for p in params.values())
if has_variadic_kwargs:
# if the signature takes **kwargs, assume they will be passed down with `super().__init__(**kwargs)`
if was_wrapped:
# if the dataloader was wrapped in a hook, only take arguments with default values
# and assume user passes their kwargs correctly
params.update({
k: v for k, v in inspect.signature(DataLoader.__init__).parameters.items() if v.default is not v.empty
})
else:
params.update(inspect.signature(DataLoader.__init__).parameters)
params.pop("self", None)
if not was_wrapped:
# keep only the params whose default is different to the current attr value
non_defaults = {name for name, p in params.items() if name in attrs and p.default is not attrs[name]}
# add `dataset` as it might have been replaced with `*args`
non_defaults.add("dataset")
# kwargs to re-construct the dataloader
dl_kwargs = {k: v for k, v in attrs.items() if k in non_defaults}
dl_args = ()
dataset = dl_kwargs.get("dataset", original_dataset)
if isinstance(dataset, IterableDataset):
dl_kwargs["batch_sampler"] = None
dl_kwargs["sampler"] = None
else:
dl_kwargs.update(_dataloader_init_kwargs_resolve_sampler(dataloader, sampler, mode))
required_args = {
p.name
for p in params.values()
if p.kind in (p.POSITIONAL_ONLY, p.POSITIONAL_OR_KEYWORD)
and p.default is p.empty
and p.name not in dl_kwargs
and p.name not in arg_names
}
# the dataloader has required args which we could not extract from the existing attributes
if required_args:
sorted_required_args = sorted(required_args)
dataloader_cls_name = dataloader.__class__.__name__
missing_args_message = ", ".join(f"`self.{arg_name}`" for arg_name in sorted_required_args)
raise MisconfigurationException(
f"Trying to inject custom `Sampler` into the `{dataloader_cls_name}` instance. "
"This would fail as some of the `__init__` arguments are not available as instance attributes. "
f"The missing attributes are {sorted_required_args}. If you instantiate your `{dataloader_cls_name}` "
"inside a `*_dataloader` hook of your module, we will do this for you."
f" Otherwise, define {missing_args_message} inside your `__init__`."
)
if not has_variadic_kwargs:
# the dataloader signature does not allow keyword arguments that need to be passed
missing_kwargs = (set(dl_kwargs) | set(arg_names)) - params.keys()
if missing_kwargs:
sorted_missing_kwargs = sorted(missing_kwargs)
dataloader_cls_name = dataloader.__class__.__name__
raise MisconfigurationException(
f"Trying to inject parameters into the `{dataloader_cls_name}` instance. "
"This would fail as it doesn't expose all its attributes in the `__init__` signature. "
f"The missing arguments are {sorted_missing_kwargs}. HINT: If you wrote the `{dataloader_cls_name}` "
"class, add the `__init__` arguments or allow passing `**kwargs`"
)
return dl_args, dl_kwargs
def _dataloader_init_kwargs_resolve_sampler(
dataloader: DataLoader,
sampler: Union[Sampler, Iterable],
mode: Optional[RunningStage] = None,
) -> dict[str, Any]:
"""This function is used to handle the sampler, batch_sampler arguments associated within a DataLoader for its re-
instantiation.
If the dataloader is being used for prediction, the sampler will be wrapped into an `_IndexBatchSamplerWrapper`, so
Lightning can keep track of its indices.
"""
is_predicting = mode == RunningStage.PREDICTING
batch_sampler = getattr(dataloader, "batch_sampler")
batch_sampler_cls = type(batch_sampler)
if batch_sampler is not None and (batch_sampler_cls is not BatchSampler or is_predicting):
if hasattr(batch_sampler, "__pl_saved_args"):
# This is a PyTorch `BatchSampler` subclass for which we captured the init args
args = batch_sampler.__pl_saved_args
kwargs = batch_sampler.__pl_saved_kwargs
default_kwargs = batch_sampler.__pl_saved_default_kwargs
arg_names = batch_sampler.__pl_saved_arg_names
if is_predicting:
success, args, kwargs = _replace_value_in_saved_args(
"drop_last", False, args, kwargs, default_kwargs, arg_names
)
if not success:
rank_zero_warn(
f"Trying to inject `drop_last=False` into batch sampler since you are predicting, however "
f"it seems the class `{batch_sampler_cls.__qualname__}` does not support it. "
"Your predictions might be incomplete. To mitigate this, expose `drop_last` in "
"the `__init__` method of your custom class."
)
success, args, kwargs = _replace_value_in_saved_args(
"sampler", sampler, args, kwargs, default_kwargs, arg_names
)
if not success:
raise TypeError(
"Trying to inject a modified sampler into the batch sampler; however, it seems the class "
f"`{batch_sampler_cls.__qualname__}` does not have an argument called `sampler.` To mitigate "
"this, expose an argument `sampler` in the `__init__` method of your custom class."
)
batch_sampler = _reinstantiate_wrapped_cls(batch_sampler, *args, **kwargs)
elif hasattr(batch_sampler, "batch_size") and hasattr(batch_sampler, "drop_last"):
# This is a sampler for which we could not capture the init args, but it kinda looks like a batch sampler
# even if it does not inherit from PyTorch's interface.
try:
batch_sampler = batch_sampler_cls(
sampler,
batch_size=batch_sampler.batch_size,
drop_last=(False if is_predicting else batch_sampler.drop_last),
)
except TypeError as ex:
import re
match = re.match(r".*__init__\(\) (got multiple values)|(missing \d required)", str(ex))
if not match:
# an unexpected `TypeError`, continue failure
raise
# There could either be too few or too many arguments. Customizing the message based on this doesn't
# make much sense since our MisconfigurationException is going to be raised from the original one.
raise TypeError(
" Lightning can't inject a (distributed) sampler into your batch sampler, because it doesn't"
" subclass PyTorch's `BatchSampler`. To mitigate this, either follow the API of `BatchSampler` and"
" instantiate your custom batch sampler inside the `*_dataloader` hook of your module,"
" or set `Trainer(use_distributed_sampler=False)`. If you choose the latter, you will be"
" responsible for handling the distributed sampling within your batch sampler."
) from ex
elif is_predicting:
rank_zero_warn(
f"You are using a custom batch sampler `{batch_sampler_cls.__qualname__}` for prediction."
" Lightning would normally set `drop_last=False` to ensure all samples are returned, but for"
" custom samplers it can't guarantee this. Make sure your sampler is configured correctly to return"
" all indices.",
category=PossibleUserWarning,
)
else:
# The sampler is not a PyTorch `BatchSampler`, we don't know how to inject a custom sampler or
# how to adjust the `drop_last` value
raise TypeError(
" Lightning can't inject a (distributed) sampler into your batch sampler, because it doesn't"
" subclass PyTorch's `BatchSampler`. To mitigate this, either follow the API of `BatchSampler`"
" or set `Trainer(use_distributed_sampler=False)`. If you choose the latter, you will be"
" responsible for handling the distributed sampling within your batch sampler."
)
if is_predicting:
batch_sampler = _IndexBatchSamplerWrapper(batch_sampler)
# batch_sampler option is mutually exclusive with batch_size, shuffle, sampler, and drop_last
return {
"sampler": None,
"shuffle": False,
"batch_sampler": batch_sampler,
"batch_size": 1,
"drop_last": False,
}
return {"sampler": sampler, "shuffle": False, "batch_sampler": None}
def _is_dataloader_shuffled(dataloader: object) -> bool:
if hasattr(dataloader, "__pl_saved_kwargs"):
# this attribute is not part of PyTorch's DataLoader, but could have been set by
# our `_replace_init_method` context manager
if "shuffle" in dataloader.__pl_saved_kwargs:
return dataloader.__pl_saved_kwargs["shuffle"]
if "shuffle" in dataloader.__pl_saved_arg_names:
return dataloader.__pl_saved_args[dataloader.__pl_saved_arg_names.index("shuffle")]
if hasattr(dataloader, "dataset") and isinstance(dataloader.dataset, IterableDataset):
# shuffling is useless with iterable datasets
return False
if not hasattr(dataloader, "sampler"):
# shuffling is enabled via a sampler. No sampler, no shuffling
return False
batch_sampler = dataloader.batch_sampler
if batch_sampler is not None:
# custom batch samplers may not have an internal .sampler
sampler = batch_sampler.sampler if hasattr(batch_sampler, "sampler") else batch_sampler
else:
sampler = dataloader.sampler
if isinstance(sampler, SequentialSampler):
return False
return isinstance(sampler, RandomSampler)
|