lightning/pytorch_lightning/trainer/connectors/checkpoint_connector.py

431 lines
18 KiB
Python

# Copyright The PyTorch Lightning 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 os
import re
from pathlib import Path
from typing import Optional, Union
import torch
import pytorch_lightning as pl
from pytorch_lightning.utilities import _OMEGACONF_AVAILABLE, rank_zero_deprecation, rank_zero_info, rank_zero_warn
from pytorch_lightning.utilities.cloud_io import atomic_save, get_filesystem
from pytorch_lightning.utilities.exceptions import MisconfigurationException
from pytorch_lightning.utilities.upgrade_checkpoint import KEYS_MAPPING as DEPRECATED_CHECKPOINT_KEYS
if _OMEGACONF_AVAILABLE:
from omegaconf import Container
class CheckpointConnector:
def __init__(self, trainer, resume_from_checkpoint: Optional[Union[str, Path]] = None):
self.trainer = trainer
self.resume_checkpoint_path = resume_from_checkpoint
self._loaded_checkpoint = {}
@property
def hpc_resume_path(self) -> Optional[str]:
dir_path_hpc = str(self.trainer.weights_save_path)
max_version = self.max_ckpt_version_in_folder(dir_path_hpc, "hpc_ckpt_")
if max_version is not None:
return os.path.join(dir_path_hpc, f"hpc_ckpt_{max_version}.ckpt")
def resume_start(self) -> None:
"""
Attempts to pre-load the checkpoint file to memory, with the source path determined in this priority:
1. from HPC weights if found
2. from `resume_from_checkpoint` file if provided
3. don't restore
Raises:
FileNotFoundError: If the path to the checkpoint file is provided but the file does not exist.
"""
self.resume_checkpoint_path = self.hpc_resume_path or self.resume_checkpoint_path
checkpoint_path = self.resume_checkpoint_path
if not checkpoint_path:
return
# clear cache before restore
torch.cuda.empty_cache()
# Try to read the checkpoint file at `checkpoint_path`. If not exist, do not restore checkpoint.
fs = get_filesystem(checkpoint_path)
if not fs.exists(checkpoint_path):
raise FileNotFoundError(f"Checkpoint at {checkpoint_path} not found. Aborting training.")
rank_zero_info(f"Restoring states from the checkpoint file at {checkpoint_path}")
self._loaded_checkpoint = self.trainer.training_type_plugin.load_checkpoint_file(checkpoint_path)
def resume_end(self) -> None:
""" Signal the connector that all states have resumed and memory for the checkpoint object can be released. """
if self.resume_checkpoint_path:
rank_zero_info(f"Restored all states from the checkpoint file at {self.resume_checkpoint_path}")
self.resume_checkpoint_path = None
self._loaded_checkpoint = {}
# clear cache after restore
torch.cuda.empty_cache()
# wait for all to catch up
self.trainer.training_type_plugin.barrier("CheckpointConnector.resume_end")
def restore(self, checkpoint_path: Optional[Union[Path, str]] = None) -> None:
"""
Attempt to restore everything at once from a 'PyTorch-Lightning checkpoint' file
through file-read and state-restore, in this priority:
1. from HPC weights if found
2. from `resume_from_checkpoint` file if provided
3. don't restore
All restored states are listed in return value description of `dump_checkpoint`.
Args:
checkpoint_path: Path to a PyTorch Lightning checkpoint file.
"""
self.resume_checkpoint_path = checkpoint_path
self.resume_start()
# restore module states
self.restore_datamodule()
self.restore_model()
# restore callback states
self.restore_callbacks()
# restore training state
self.restore_training_state()
self.resume_end()
def restore_datamodule(self) -> None:
""" Calls hooks on the datamodule to give it a chance to restore its state from the checkpoint. """
if not self._loaded_checkpoint:
return
datamodule = self.trainer.datamodule
if datamodule is not None:
datamodule.on_load_checkpoint(self._loaded_checkpoint)
def restore_model(self) -> None:
"""
Restores a model's weights from a PyTorch Lightning checkpoint. Hooks are called first go give
the LightningModule a chance to modify the contents, then finally the model gets updated with
the loaded weights.
"""
if not self._loaded_checkpoint:
return
model = self.trainer.lightning_module
# hook: give user access to checkpoint if needed.
model.on_load_checkpoint(self._loaded_checkpoint)
# call hpc specific hook
if self.hpc_resume_path is not None:
model.on_hpc_load(self._loaded_checkpoint)
# restore model state_dict
self.trainer.training_type_plugin.load_model_state_dict(self._loaded_checkpoint)
def restore_model_weights(self, checkpoint_path: Optional[Union[str, Path]]) -> None:
""" Restore only the model weights. """
checkpoint = self._loaded_checkpoint
if checkpoint_path is not None:
checkpoint = self.trainer.training_type_plugin.load_checkpoint_file(checkpoint_path)
self.trainer.lightning_module.on_load_checkpoint(checkpoint)
self.trainer.training_type_plugin.load_model_state_dict(checkpoint)
def restore_training_state(self) -> None:
"""
Restore the trainer state from the pre-loaded checkpoint. This includes the precision settings, loop progress,
optimizer states and learning rate scheduler states.
"""
if not self._loaded_checkpoint:
return
# restore precision plugin (scaler etc.)
self.trainer.precision_plugin.on_load_checkpoint(self._loaded_checkpoint)
# restore progress (loops etc.)
self.restore_progress()
self.restore_optimizers_and_schedulers()
def restore_callbacks(self) -> None:
""" Restores all callbacks from the pre-loaded checkpoint. """
if not self._loaded_checkpoint:
return
if any(key in self._loaded_checkpoint for key in DEPRECATED_CHECKPOINT_KEYS):
raise ValueError(
"The checkpoint you're attempting to load follows an"
" outdated schema. You can upgrade to the current schema by running"
" `python -m pytorch_lightning.utilities.upgrade_checkpoint --file model.ckpt`"
" where `model.ckpt` is your checkpoint file."
)
self.trainer.on_load_checkpoint(self._loaded_checkpoint)
def restore_progress(self) -> None:
"""
Restores the training progress from the pre-loaded checkpoint. This currently includes only the global step
and current epoch.
"""
if not self._loaded_checkpoint:
return
self.trainer.fit_loop.global_step = self._loaded_checkpoint['global_step']
self.trainer.fit_loop.current_epoch = self._loaded_checkpoint['epoch']
# crash if max_epochs is lower then the current epoch from the checkpoint
if self.trainer.max_epochs is not None and self.trainer.current_epoch > self.trainer.max_epochs:
raise MisconfigurationException(
f"You restored a checkpoint with current_epoch={self.trainer.current_epoch},"
f" but you have set Trainer(max_epochs={self.trainer.max_epochs})."
)
# Division deals with global step stepping once per accumulated batch
# Inequality deals with different global step for odd vs even num_training_batches
n_accum = 1 if self.trainer.accumulate_grad_batches is None else self.trainer.accumulate_grad_batches
expected_steps = self.trainer.num_training_batches / n_accum
if self.trainer.num_training_batches != 0 and self.trainer.global_step % expected_steps > 1:
rank_zero_warn(
"You're resuming from a checkpoint that ended mid-epoch."
" Training will start from the beginning of the next epoch."
" This can cause unreliable results if further training is done,"
" consider using an end of epoch checkpoint."
)
def restore_optimizers_and_schedulers(self) -> None:
""" Restores the optimizers and learning rate scheduler states from the pre-loaded checkpoint. """
if not self._loaded_checkpoint:
return
# validation
if "optimizer_states" not in self._loaded_checkpoint or "lr_schedulers" not in self._loaded_checkpoint:
raise KeyError(
"Trying to restore training state but checkpoint contains only the model."
" This is probably due to `ModelCheckpoint.save_weights_only` being set to `True`."
)
self.restore_optimizers()
self.restore_lr_schedulers()
def restore_optimizers(self) -> None:
""" Restores the optimizer states from the pre-loaded checkpoint. """
if not self._loaded_checkpoint:
return
# restore the optimizers
self.trainer.training_type_plugin.load_optimizer_state_dict(self._loaded_checkpoint)
for optimizer in self.trainer.optimizers:
# move optimizer to GPU 1 weight at a time
# avoids OOM
if self.trainer.root_gpu is not None:
for state in optimizer.state.values():
for k, v in state.items():
if isinstance(v, torch.Tensor):
state[k] = v.cuda(self.trainer.root_gpu)
def restore_lr_schedulers(self) -> None:
""" Restores the learning rate scheduler states from the pre-loaded checkpoint. """
if not self._loaded_checkpoint:
return
# restore the lr schedulers
lr_schedulers = self._loaded_checkpoint['lr_schedulers']
for scheduler, lrs_state in zip(self.trainer.lr_schedulers, lr_schedulers):
scheduler['scheduler'].load_state_dict(lrs_state)
# ----------------------------------
# PRIVATE OPS
# ----------------------------------
def hpc_save(self, folderpath: str, logger):
# make sure the checkpoint folder exists
folderpath = str(folderpath) # because the tests pass a path object
fs = get_filesystem(folderpath)
fs.makedirs(folderpath, exist_ok=True)
# save logger to make sure we get all the metrics
logger.save()
max_suffix = self.max_ckpt_version_in_folder(folderpath)
ckpt_number = (max_suffix if max_suffix is not None else 0) + 1
fs.makedirs(folderpath, exist_ok=True)
filepath = os.path.join(folderpath, f'hpc_ckpt_{ckpt_number}.ckpt')
# give model a chance to do something on hpc_save
model = self.trainer.lightning_module
checkpoint = self.dump_checkpoint()
model.on_hpc_save(checkpoint)
checkpoint = self.trainer.accelerator.on_save(checkpoint)
# do the actual save
# TODO: fix for anything with multiprocess DP, DDP, DDP2
try:
atomic_save(checkpoint, filepath)
except AttributeError as err:
if pl.LightningModule.CHECKPOINT_HYPER_PARAMS_KEY in checkpoint:
del checkpoint[pl.LightningModule.CHECKPOINT_HYPER_PARAMS_KEY]
rank_zero_warn(
'warning, `hyper_parameters` dropped from checkpoint.'
f' An attribute is not picklable {err}'
)
atomic_save(checkpoint, filepath)
return filepath
def dump_checkpoint(self, weights_only: bool = False) -> dict:
"""Creating a model checkpoint dictionary object from various component states.
Args:
weights_only: saving model weights only
Return:
structured dictionary: {
'epoch': training epoch
'global_step': training global step
'pytorch-lightning_version': PyTorch Lightning's version
'callbacks': "callback specific state"[] # if not weights_only
'optimizer_states': "PT optim's state_dict"[] # if not weights_only
'lr_schedulers': "PT sched's state_dict"[] # if not weights_only
'native_amp_scaling_state': PT amp's state_dict # if not weights_only and use native amp
'amp_scaling_state': Apex's state_dict # if not weights_only and use apex amp
'state_dict': Model's state_dict (e.g. network weights)
CHECKPOINT_HYPER_PARAMS_NAME:
CHECKPOINT_HYPER_PARAMS_KEY:
CHECKPOINT_HYPER_PARAMS_TYPE:
something_cool_i_want_to_save: anything you define through model.on_save_checkpoint
LightningDataModule.__class__.__name__: pl DataModule's state
}
"""
# dump epoch/global_step/pytorch-lightning_version
current_epoch = self.trainer.current_epoch
global_step = self.trainer.global_step
has_reached_max_steps = self.trainer.max_steps and self.trainer.max_steps <= global_step
global_step += 1
if not has_reached_max_steps:
current_epoch += 1
model = self.trainer.lightning_module
checkpoint = {
'epoch': current_epoch,
'global_step': global_step,
'pytorch-lightning_version': pl.__version__,
'state_dict': self.trainer.accelerator.lightning_module_state_dict(),
}
if not weights_only:
# dump callbacks
checkpoint['callbacks'] = self.trainer.on_save_checkpoint(checkpoint)
optimizer_states = []
for i, optimizer in enumerate(self.trainer.optimizers):
# Rely on accelerator to dump optimizer state
optimizer_state = self.trainer.accelerator.optimizer_state(optimizer)
optimizer_states.append(optimizer_state)
checkpoint['optimizer_states'] = optimizer_states
# dump lr schedulers
lr_schedulers = []
for scheduler in self.trainer.lr_schedulers:
lr_schedulers.append(scheduler['scheduler'].state_dict())
checkpoint['lr_schedulers'] = lr_schedulers
self.trainer.precision_plugin.on_save_checkpoint(checkpoint)
# dump hyper-parameters
if model.hparams:
if hasattr(model, '_hparams_name'):
checkpoint[pl.LightningModule.CHECKPOINT_HYPER_PARAMS_NAME] = model._hparams_name
# dump arguments
if _OMEGACONF_AVAILABLE and isinstance(model.hparams, Container):
checkpoint[pl.LightningModule.CHECKPOINT_HYPER_PARAMS_KEY] = model.hparams
checkpoint[pl.LightningModule.CHECKPOINT_HYPER_PARAMS_TYPE] = type(model.hparams)
else:
checkpoint[pl.LightningModule.CHECKPOINT_HYPER_PARAMS_KEY] = dict(model.hparams)
# give the model a chance to dump a few things
model.on_save_checkpoint(checkpoint)
if self.trainer.datamodule is not None:
self.trainer.datamodule.on_save_checkpoint(checkpoint)
return checkpoint
def hpc_load(self, checkpoint_path: str) -> None:
"""
Attempts to restore the full training and model state from a HPC checkpoint file.
.. deprecated::v1.4
Will be removed in v1.6. Use :meth:`restore` instead.
"""
rank_zero_deprecation(
"`CheckpointConnector.hpc_load()` was deprecated in v1.4 and will be removed in v1.6."
" Use `CheckpointConnector.restore()` instead."
)
self.restore(checkpoint_path)
def max_ckpt_version_in_folder(self, dir_path: Union[str, Path], name_key: str = 'ckpt_') -> Optional[int]:
"""List up files in `dir_path` with `name_key`, then yield maximum suffix number.
Args:
dir_path: path of directory which may contain files whose name include `name_key`
name_key: file name prefix
Returns:
None if no-corresponding-file else maximum suffix number
"""
# check directory existence
fs = get_filesystem(dir_path)
if not fs.exists(dir_path):
return None
# check corresponding file existence
files = [os.path.basename(f["name"]) for f in fs.listdir(dir_path)]
files = [x for x in files if name_key in x]
if len(files) == 0:
return None
# extract suffix number
ckpt_vs = []
for name in files:
name = name.split(name_key)[-1]
name = re.sub('[^0-9]', '', name)
ckpt_vs.append(int(name))
return max(ckpt_vs)
def get_max_ckpt_path_from_folder(self, folder_path: Union[str, Path]) -> str:
"""Get path of maximum-epoch checkpoint in the folder."""
max_suffix = self.max_ckpt_version_in_folder(folder_path)
ckpt_number = max_suffix if max_suffix is not None else 0
return f'{folder_path}/hpc_ckpt_{ckpt_number}.ckpt'
def save_checkpoint(self, filepath, weights_only: bool = False) -> None:
"""Save model/training states as a checkpoint file through state-dump and file-write.
Args:
filepath: write-target file's path
weights_only: saving model weights only
"""
_checkpoint = self.dump_checkpoint(weights_only)
self.trainer.accelerator.save_checkpoint(_checkpoint, filepath)