.. role:: hidden
:class: hidden-section
.. testsetup:: *
import os
from pytorch_lightning.trainer.trainer import Trainer
from pytorch_lightning.core.lightning import LightningModule
from pytorch_lightning.utilities.seed import seed_everything
.. _trainer:
Trainer
=======
Once you've organized your PyTorch code into a LightningModule,
the Trainer automates everything else.
.. raw:: html
|
This abstraction achieves the following:
1. You maintain control over all aspects via PyTorch code without an added abstraction.
2. The trainer uses best practices embedded by contributors and users
from top AI labs such as Facebook AI Research, NYU, MIT, Stanford, etc...
3. The trainer allows overriding any key part that you don't want automated.
|
-----------
Basic use
---------
This is the basic use of the trainer:
.. code-block:: python
model = MyLightningModule()
trainer = Trainer()
trainer.fit(model, train_dataloader, val_dataloader)
--------
Under the hood
--------------
Under the hood, the Lightning Trainer handles the training loop details for you, some examples include:
- Automatically enabling/disabling grads
- Running the training, validation and test dataloaders
- Calling the Callbacks at the appropriate times
- Putting batches and computations on the correct devices
Here's the pseudocode for what the trainer does under the hood (showing the train loop only)
.. code-block:: python
# put model in train mode
model.train()
torch.set_grad_enabled(True)
losses = []
for batch in train_dataloader:
# calls hooks like this one
on_train_batch_start()
# train step
loss = training_step(batch)
# backward
loss.backward()
# apply and clear grads
optimizer.step()
optimizer.zero_grad()
losses.append(loss)
--------
Trainer in Python scripts
-------------------------
In Python scripts, it's recommended you use a main function to call the Trainer.
.. code-block:: python
from argparse import ArgumentParser
def main(hparams):
model = LightningModule()
trainer = Trainer(gpus=hparams.gpus)
trainer.fit(model)
if __name__ == '__main__':
parser = ArgumentParser()
parser.add_argument('--gpus', default=None)
args = parser.parse_args()
main(args)
So you can run it like so:
.. code-block:: bash
python main.py --gpus 2
.. note::
Pro-tip: You don't need to define all flags manually. Lightning can add them automatically
.. code-block:: python
from argparse import ArgumentParser
def main(args):
model = LightningModule()
trainer = Trainer.from_argparse_args(args)
trainer.fit(model)
if __name__ == '__main__':
parser = ArgumentParser()
parser = Trainer.add_argparse_args(parser)
args = parser.parse_args()
main(args)
So you can run it like so:
.. code-block:: bash
python main.py --gpus 2 --max_steps 10 --limit_train_batches 10 --any_trainer_arg x
.. note::
If you want to stop a training run early, you can press "Ctrl + C" on your keyboard.
The trainer will catch the ``KeyboardInterrupt`` and attempt a graceful shutdown, including
running callbacks such as ``on_train_end``. The trainer object will also set an attribute
``interrupted`` to ``True`` in such cases. If you have a callback which shuts down compute
resources, for example, you can conditionally run the shutdown logic for only uninterrupted runs.
------------
Testing
-------
Once you're done training, feel free to run the test set!
(Only right before publishing your paper or pushing to production)
.. code-block:: python
trainer.test(test_dataloaders=test_dataloader)
------------
Deployment / prediction
-----------------------
You just trained a LightningModule which is also just a torch.nn.Module.
Use it to do whatever!
.. code-block:: python
# load model
pretrained_model = LightningModule.load_from_checkpoint(PATH)
pretrained_model.freeze()
# use it for finetuning
def forward(self, x):
features = pretrained_model(x)
classes = classifier(features)
# or for prediction
out = pretrained_model(x)
api_write({'response': out}
You may wish to run the model on a variety of devices. Instead of moving the data
manually to the correct device, decorate the forward method (or any other method you use for inference)
with :func:`~pytorch_lightning.core.decorators.auto_move_data` and Lightning will take care of the rest.
------------
Reproducibility
---------------
To ensure full reproducibility from run to run you need to set seeds for pseudo-random generators,
and set ``deterministic`` flag in ``Trainer``.
Example::
from pytorch_lightning import Trainer, seed_everything
seed_everything(42)
# sets seeds for numpy, torch, python.random and PYTHONHASHSEED.
model = Model()
trainer = Trainer(deterministic=True)
-------
Trainer flags
-------------
accelerator
^^^^^^^^^^^
.. raw:: html
|
The accelerator backend to use (previously known as distributed_backend).
- (``'dp'``) is DataParallel (split batch among GPUs of same machine)
- (``'ddp'``) is DistributedDataParallel (each gpu on each node trains, and syncs grads)
- (``'ddp_cpu'``) is DistributedDataParallel on CPU (same as ``'ddp'``, but does not use GPUs.
Useful for multi-node CPU training or single-node debugging. Note that this will **not** give
a speedup on a single node, since Torch already makes efficient use of multiple CPUs on a single
machine.)
- (``'ddp2'``) dp on node, ddp across nodes. Useful for things like increasing
the number of negative samples
.. testcode::
# default used by the Trainer
trainer = Trainer(accelerator=None)
Example::
# dp = DataParallel
trainer = Trainer(gpus=2, accelerator='dp')
# ddp = DistributedDataParallel
trainer = Trainer(gpus=2, num_nodes=2, accelerator='ddp')
# ddp2 = DistributedDataParallel + dp
trainer = Trainer(gpus=2, num_nodes=2, accelerator='ddp2')
.. note:: This option does not apply to TPU. TPUs use ``'ddp'`` by default (over each core)
You can also modify hardware behavior by subclassing an existing accelerator to adjust for your needs.
Example::
class MyOwnDDP(DDPAccelerator):
...
Trainer(accelerator=MyOwnDDP())
.. warning:: Passing in custom accelerators is experimental but work is in progress to enable full compatibility.
accumulate_grad_batches
^^^^^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
Accumulates grads every k batches or as set up in the dict.
Trainer also calls ``optimizer.step()`` for the last indivisible step number.
.. testcode::
# default used by the Trainer (no accumulation)
trainer = Trainer(accumulate_grad_batches=1)
Example::
# accumulate every 4 batches (effective batch size is batch*4)
trainer = Trainer(accumulate_grad_batches=4)
# no accumulation for epochs 1-4. accumulate 3 for epochs 5-10. accumulate 20 after that
trainer = Trainer(accumulate_grad_batches={5: 3, 10: 20})
amp_backend
^^^^^^^^^^^
.. raw:: html
|
Use PyTorch AMP ('native') (available PyTorch 1.6+), or NVIDIA apex ('apex').
.. testcode::
# using PyTorch built-in AMP, default used by the Trainer
trainer = Trainer(amp_backend='native')
# using NVIDIA Apex
trainer = Trainer(amp_backend='apex')
amp_level
^^^^^^^^^
.. raw:: html
|
The optimization level to use (O1, O2, etc...)
for 16-bit GPU precision (using NVIDIA apex under the hood).
Check `NVIDIA apex docs `_ for level
Example::
# default used by the Trainer
trainer = Trainer(amp_level='O2')
automatic_optimization
^^^^^^^^^^^^^^^^^^^^^^
When set to False, Lightning does not automate the optimization process. This means you are responsible for your own
optimizer behavior
Example::
def training_step(self, batch, batch_idx):
# access your optimizers with use_pl_optimizer=False. Default is True
opt = self.optimizers(use_pl_optimizer=True)
loss = ...
self.manual_backward(loss, opt)
opt.step()
opt.zero_grad()
This is not recommended when using a single optimizer, instead it's recommended when using 2+ optimizers
AND you are an expert user. Most useful for research like RL, sparse coding and GAN research.
In the multi-optimizer case, ignore the optimizer_idx flag and use the optimizers directly
Example::
def training_step(self, batch, batch_idx, optimizer_idx):
# access your optimizers with use_pl_optimizer=False. Default is True
(opt_a, opt_b) = self.optimizers(use_pl_optimizer=True)
gen_loss = ...
self.manual_backward(gen_loss, opt_a)
opt_a.step()
opt_a.zero_grad()
disc_loss = ...
self.manual_backward(disc_loss, opt_b)
opt_b.step()
opt_b.zero_grad()
auto_scale_batch_size
^^^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
Automatically tries to find the largest batch size that fits into memory,
before any training.
.. code-block::
# default used by the Trainer (no scaling of batch size)
trainer = Trainer(auto_scale_batch_size=None)
# run batch size scaling, result overrides hparams.batch_size
trainer = Trainer(auto_scale_batch_size='binsearch')
# call tune to find the batch size
trainer.tune(model)
auto_select_gpus
^^^^^^^^^^^^^^^^
.. raw:: html
|
If enabled and `gpus` is an integer, pick available gpus automatically.
This is especially useful when GPUs are configured to be in "exclusive mode",
such that only one process at a time can access them.
Example::
# no auto selection (picks first 2 gpus on system, may fail if other process is occupying)
trainer = Trainer(gpus=2, auto_select_gpus=False)
# enable auto selection (will find two available gpus on system)
trainer = Trainer(gpus=2, auto_select_gpus=True)
# specifies all GPUs regardless of its availability
Trainer(gpus=-1, auto_select_gpus=False)
# specifies all available GPUs (if only one GPU is not occupied, uses one gpu)
Trainer(gpus=-1, auto_select_gpus=True)
auto_lr_find
^^^^^^^^^^^^
.. raw:: html
|
Runs a learning rate finder algorithm (see this `paper `_)
when calling trainer.tune(), to find optimal initial learning rate.
.. code-block:: python
# default used by the Trainer (no learning rate finder)
trainer = Trainer(auto_lr_find=False)
Example::
# run learning rate finder, results override hparams.learning_rate
trainer = Trainer(auto_lr_find=True)
# call tune to find the lr
trainer.tune(model)
Example::
# run learning rate finder, results override hparams.my_lr_arg
trainer = Trainer(auto_lr_find='my_lr_arg')
# call tune to find the lr
trainer.tune(model)
.. note::
See the :ref:`learning rate finder guide `.
benchmark
^^^^^^^^^
.. raw:: html
|
If true enables cudnn.benchmark.
This flag is likely to increase the speed of your system if your
input sizes don't change. However, if it does, then it will likely
make your system slower.
The speedup comes from allowing the cudnn auto-tuner to find the best
algorithm for the hardware `[see discussion here]
`_.
Example::
# default used by the Trainer
trainer = Trainer(benchmark=False)
deterministic
^^^^^^^^^^^^^
.. raw:: html
|
If true enables cudnn.deterministic.
Might make your system slower, but ensures reproducibility.
Also sets ``$HOROVOD_FUSION_THRESHOLD=0``.
For more info check `[pytorch docs]
`_.
Example::
# default used by the Trainer
trainer = Trainer(deterministic=False)
callbacks
^^^^^^^^^
.. raw:: html
|
Add a list of :class:`~pytorch_lightning.callbacks.Callback`.
.. code-block:: python
# a list of callbacks
callbacks = [PrintCallback()]
trainer = Trainer(callbacks=callbacks)
Example::
from pytorch_lightning.callbacks import Callback
class PrintCallback(Callback):
def on_train_start(self, trainer, pl_module):
print("Training is started!")
def on_train_end(self, trainer, pl_module):
print("Training is done.")
check_val_every_n_epoch
^^^^^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
Check val every n train epochs.
Example::
# default used by the Trainer
trainer = Trainer(check_val_every_n_epoch=1)
# run val loop every 10 training epochs
trainer = Trainer(check_val_every_n_epoch=10)
checkpoint_callback
^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
By default Lightning saves a checkpoint for you in your current working directory, with the state of your last training epoch,
Checkpoints capture the exact value of all parameters used by a model.
To disable automatic checkpointing, set this to `False`.
.. code-block:: python
# default used by Trainer
trainer = Trainer(checkpoint_callback=True)
# turn off automatic checkpointing
trainer = Trainer(checkpoint_callback=False)
You can override the default behavior by initializing the :class:`~pytorch_lightning.callbacks.ModelCheckpoint`
callback, and adding it to the :paramref:`~pytorch_lightning.trainer.trainer.Trainer.callbacks` list.
See :ref:`Saving and Loading Weights ` for how to customize checkpointing.
.. testcode::
from pytorch_lightning.callbacks import ModelCheckpoint
# Init ModelCheckpoint callback, monitoring 'val_loss'
checkpoint_callback = ModelCheckpoint(monitor='val_loss')
# Add your callback to the callbacks list
trainer = Trainer(callbacks=[checkpoint_callback])
.. warning:: Passing a ModelCheckpoint instance to this argument is deprecated since
v1.1 and will be unsupported from v1.3. Use `callbacks` argument instead.
default_root_dir
^^^^^^^^^^^^^^^^
.. raw:: html
|
Default path for logs and weights when no logger or
:class:`pytorch_lightning.callbacks.ModelCheckpoint` callback passed. On
certain clusters you might want to separate where logs and checkpoints are
stored. If you don't then use this argument for convenience. Paths can be local
paths or remote paths such as `s3://bucket/path` or 'hdfs://path/'. Credentials
will need to be set up to use remote filepaths.
.. testcode::
# default used by the Trainer
trainer = Trainer(default_root_dir=os.getcwd())
distributed_backend
^^^^^^^^^^^^^^^^^^^
Deprecated: This has been renamed ``accelerator``.
fast_dev_run
^^^^^^^^^^^^
.. raw:: html
|
Runs n if set to ``n`` (int) else 1 if set to ``True`` batch(es) of train, val and test
to find any bugs (ie: a sort of unit test).
Under the hood the pseudocode looks like this when running *fast_dev_run* with a single batch:
.. code-block:: python
# loading
__init__()
prepare_data
# test training step
training_batch = next(train_dataloader)
training_step(training_batch)
# test val step
val_batch = next(val_dataloader)
out = validation_step(val_batch)
validation_epoch_end([out])
.. testcode::
# default used by the Trainer
trainer = Trainer(fast_dev_run=False)
# runs 1 train, val, test batch and program ends
trainer = Trainer(fast_dev_run=True)
# runs 7 train, val, test batches and program ends
trainer = Trainer(fast_dev_run=7)
.. note::
This argument is a bit different from ``limit_train/val/test_batches``. Setting this argument will
disable tuner, checkpoint callbacks, early stopping callbacks, loggers and logger callbacks like
``LearningRateLogger`` and runs for only 1 epoch. This must be used only for debugging purposes.
``limit_train/val/test_batches`` only limits the number of batches and won't disable anything.
flush_logs_every_n_steps
^^^^^^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
Writes logs to disk this often.
.. testcode::
# default used by the Trainer
trainer = Trainer(flush_logs_every_n_steps=100)
See Also:
- :ref:`logging`
gpus
^^^^
.. raw:: html
|
- Number of GPUs to train on (int)
- or which GPUs to train on (list)
- can handle strings
.. testcode::
# default used by the Trainer (ie: train on CPU)
trainer = Trainer(gpus=None)
# equivalent
trainer = Trainer(gpus=0)
Example::
# int: train on 2 gpus
trainer = Trainer(gpus=2)
# list: train on GPUs 1, 4 (by bus ordering)
trainer = Trainer(gpus=[1, 4])
trainer = Trainer(gpus='1, 4') # equivalent
# -1: train on all gpus
trainer = Trainer(gpus=-1)
trainer = Trainer(gpus='-1') # equivalent
# combine with num_nodes to train on multiple GPUs across nodes
# uses 8 gpus in total
trainer = Trainer(gpus=2, num_nodes=4)
# train only on GPUs 1 and 4 across nodes
trainer = Trainer(gpus=[1, 4], num_nodes=4)
See Also:
- :ref:`Multi-GPU training guide `.
gradient_clip_val
^^^^^^^^^^^^^^^^^
.. raw:: html
|
Gradient clipping value
- 0 means don't clip.
.. testcode::
# default used by the Trainer
trainer = Trainer(gradient_clip_val=0.0)
limit_train_batches
^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
How much of training dataset to check.
Useful when debugging or testing something that happens at the end of an epoch.
.. testcode::
# default used by the Trainer
trainer = Trainer(limit_train_batches=1.0)
Example::
# default used by the Trainer
trainer = Trainer(limit_train_batches=1.0)
# run through only 25% of the training set each epoch
trainer = Trainer(limit_train_batches=0.25)
# run through only 10 batches of the training set each epoch
trainer = Trainer(limit_train_batches=10)
limit_test_batches
^^^^^^^^^^^^^^^^^^
.. raw:: html
|
How much of test dataset to check.
.. testcode::
# default used by the Trainer
trainer = Trainer(limit_test_batches=1.0)
# run through only 25% of the test set each epoch
trainer = Trainer(limit_test_batches=0.25)
# run for only 10 batches
trainer = Trainer(limit_test_batches=10)
In the case of multiple test dataloaders, the limit applies to each dataloader individually.
limit_val_batches
^^^^^^^^^^^^^^^^^
.. raw:: html
|
How much of validation dataset to check.
Useful when debugging or testing something that happens at the end of an epoch.
.. testcode::
# default used by the Trainer
trainer = Trainer(limit_val_batches=1.0)
# run through only 25% of the validation set each epoch
trainer = Trainer(limit_val_batches=0.25)
# run for only 10 batches
trainer = Trainer(limit_val_batches=10)
In the case of multiple validation dataloaders, the limit applies to each dataloader individually.
log_every_n_steps
^^^^^^^^^^^^^^^^^
.. raw:: html
|
How often to add logging rows (does not write to disk)
.. testcode::
# default used by the Trainer
trainer = Trainer(log_every_n_steps=50)
See Also:
- :ref:`logging`
log_gpu_memory
^^^^^^^^^^^^^^
.. raw:: html
|
Options:
- None
- 'min_max'
- 'all'
.. testcode::
# default used by the Trainer
trainer = Trainer(log_gpu_memory=None)
# log all the GPUs (on master node only)
trainer = Trainer(log_gpu_memory='all')
# log only the min and max memory on the master node
trainer = Trainer(log_gpu_memory='min_max')
.. note:: Might slow performance because it uses the output of ``nvidia-smi``.
logger
^^^^^^
.. raw:: html
|
:ref:`Logger ` (or iterable collection of loggers) for experiment tracking.
.. testcode::
from pytorch_lightning.loggers import TensorBoardLogger
# default logger used by trainer
logger = TensorBoardLogger(
save_dir=os.getcwd(),
version=1,
name='lightning_logs'
)
Trainer(logger=logger)
max_epochs
^^^^^^^^^^
.. raw:: html
|
Stop training once this number of epochs is reached
.. testcode::
# default used by the Trainer
trainer = Trainer(max_epochs=1000)
min_epochs
^^^^^^^^^^
.. raw:: html
|
Force training for at least these many epochs
.. testcode::
# default used by the Trainer
trainer = Trainer(min_epochs=1)
max_steps
^^^^^^^^^
.. raw:: html
|
Stop training after this number of steps
Training will stop if max_steps or max_epochs have reached (earliest).
.. testcode::
# Default (disabled)
trainer = Trainer(max_steps=None)
# Stop after 100 steps
trainer = Trainer(max_steps=100)
min_steps
^^^^^^^^^
.. raw:: html
|
Force training for at least these number of steps.
Trainer will train model for at least min_steps or min_epochs (latest).
.. testcode::
# Default (disabled)
trainer = Trainer(min_steps=None)
# Run at least for 100 steps (disable min_epochs)
trainer = Trainer(min_steps=100, min_epochs=0)
num_nodes
^^^^^^^^^
.. raw:: html
|
Number of GPU nodes for distributed training.
.. testcode::
# default used by the Trainer
trainer = Trainer(num_nodes=1)
# to train on 8 nodes
trainer = Trainer(num_nodes=8)
num_processes
^^^^^^^^^^^^^
.. raw:: html
|
Number of processes to train with. Automatically set to the number of GPUs
when using ``accelerator="ddp"``. Set to a number greater than 1 when
using ``accelerator="ddp_cpu"`` to mimic distributed training on a
machine without GPUs. This is useful for debugging, but **will not** provide
any speedup, since single-process Torch already makes efficient use of multiple
CPUs.
.. testcode::
# Simulate DDP for debugging on your GPU-less laptop
trainer = Trainer(accelerator="ddp_cpu", num_processes=2)
num_sanity_val_steps
^^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
Sanity check runs n batches of val before starting the training routine.
This catches any bugs in your validation without having to wait for the first validation check.
The Trainer uses 2 steps by default. Turn it off or modify it here.
.. testcode::
# default used by the Trainer
trainer = Trainer(num_sanity_val_steps=2)
# turn it off
trainer = Trainer(num_sanity_val_steps=0)
# check all validation data
trainer = Trainer(num_sanity_val_steps=-1)
This option will reset the validation dataloader unless ``num_sanity_val_steps=0``.
overfit_batches
^^^^^^^^^^^^^^^
.. raw:: html
|
Uses this much data of the training set. If nonzero, will use the same training set for validation and testing.
If the training dataloaders have `shuffle=True`, Lightning will automatically disable it.
Useful for quickly debugging or trying to overfit on purpose.
.. testcode::
# default used by the Trainer
trainer = Trainer(overfit_batches=0.0)
# use only 1% of the train set (and use the train set for val and test)
trainer = Trainer(overfit_batches=0.01)
# overfit on 10 of the same batches
trainer = Trainer(overfit_batches=10)
plugins
^^^^^^^
.. raw:: html
|
Plugins allow you to connect arbitrary backends, precision libraries, SLURM, etc... For example:
- DDP
- SLURM
- TorchElastic
- Apex
To define your own behavior, subclass the relevant class and pass it in. Here's an example linking up your own cluster.
.. code-block:: python
from pytorch_lightning.cluster_environments import cluster_environment
class MyCluster(ClusterEnvironment):
def master_address(self):
return your_master_address
def master_port(self):
return your_master_port
def world_size(self):
return the_world_size
trainer = Trainer(cluster_environment=cluster_environment())
prepare_data_per_node
^^^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
If True will call `prepare_data()` on LOCAL_RANK=0 for every node.
If False will only call from NODE_RANK=0, LOCAL_RANK=0
.. testcode::
# default
Trainer(prepare_data_per_node=True)
# use only NODE_RANK=0, LOCAL_RANK=0
Trainer(prepare_data_per_node=False)
precision
^^^^^^^^^
.. raw:: html
|
Full precision (32), half precision (16).
Can be used on CPU, GPU or TPUs.
If used on TPU will use torch.bfloat16 but tensor printing
will still show torch.float32.
.. testcode::
:skipif: not _APEX_AVAILABLE and not _NATIVE_AMP_AVAILABLE
# default used by the Trainer
trainer = Trainer(precision=32)
# 16-bit precision
trainer = Trainer(precision=16)
Example::
# one day
trainer = Trainer(precision=8|4|2)
process_position
^^^^^^^^^^^^^^^^
.. raw:: html
|
Orders the progress bar. Useful when running multiple trainers on the same node.
.. testcode::
# default used by the Trainer
trainer = Trainer(process_position=0)
.. note:: This argument is ignored if a custom callback is passed to :paramref:`~Trainer.callbacks`.
profiler
^^^^^^^^
.. raw:: html
|
To profile individual steps during training and assist in identifying bottlenecks.
See the :ref:`profiler documentation `. for more details.
.. testcode::
from pytorch_lightning.profiler import SimpleProfiler, AdvancedProfiler
# default used by the Trainer
trainer = Trainer(profiler=None)
# to profile standard training events, equivalent to `profiler=SimpleProfiler()`
trainer = Trainer(profiler="simple")
# advanced profiler for function-level stats, equivalent to `profiler=AdvancedProfiler()`
trainer = Trainer(profiler="advanced")
progress_bar_refresh_rate
^^^^^^^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
How often to refresh progress bar (in steps).
.. testcode::
# default used by the Trainer
trainer = Trainer(progress_bar_refresh_rate=1)
# disable progress bar
trainer = Trainer(progress_bar_refresh_rate=0)
Note:
- In Google Colab notebooks, faster refresh rates (lower number) is known to crash them because of their screen refresh rates.
Lightning will set it to 20 in these environments if the user does not provide a value.
- This argument is ignored if a custom callback is passed to :paramref:`~Trainer.callbacks`.
reload_dataloaders_every_epoch
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
Set to True to reload dataloaders every epoch.
.. code-block:: python
# if False (default)
train_loader = model.train_dataloader()
for epoch in epochs:
for batch in train_loader:
...
# if True
for epoch in epochs:
train_loader = model.train_dataloader()
for batch in train_loader:
replace_sampler_ddp
^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
Enables auto adding of distributed sampler. By default it will add ``shuffle=True``
for train sampler and ``shuffle=False`` for val/test sampler. If you want to customize
it, you can set ``replace_sampler_ddp=False`` and add your own distributed sampler.
If ``replace_sampler_ddp=True`` and a distributed sampler was already added,
Lightning will not replace the existing one.
.. testcode::
# default used by the Trainer
trainer = Trainer(replace_sampler_ddp=True)
By setting to False, you have to add your own distributed sampler:
.. code-block:: python
# default used by the Trainer
sampler = torch.utils.data.distributed.DistributedSampler(dataset, shuffle=True)
dataloader = DataLoader(dataset, batch_size=32, sampler=sampler)
resume_from_checkpoint
^^^^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
To resume training from a specific checkpoint pass in the path here. If resuming from a mid-epoch
checkpoint, training will start from the beginning of the next epoch.
.. testcode::
# default used by the Trainer
trainer = Trainer(resume_from_checkpoint=None)
# resume from a specific checkpoint
trainer = Trainer(resume_from_checkpoint='some/path/to/my_checkpoint.ckpt')
sync_batchnorm
^^^^^^^^^^^^^^
.. raw:: html
|
Enable synchronization between batchnorm layers across all GPUs.
.. testcode::
trainer = Trainer(sync_batchnorm=True)
track_grad_norm
^^^^^^^^^^^^^^^
.. raw:: html
|
- no tracking (-1)
- Otherwise tracks that norm (2 for 2-norm)
.. testcode::
# default used by the Trainer
trainer = Trainer(track_grad_norm=-1)
# track the 2-norm
trainer = Trainer(track_grad_norm=2)
tpu_cores
^^^^^^^^^
.. raw:: html
|
- How many TPU cores to train on (1 or 8).
- Which TPU core to train on [1-8]
A single TPU v2 or v3 has 8 cores. A TPU pod has
up to 2048 cores. A slice of a POD means you get as many cores
as you request.
Your effective batch size is batch_size * total tpu cores.
.. note::
No need to add a :class:`~torch.utils.data.distributed.DistributedSampler`,
Lightning automatically does it for you.
This parameter can be either 1 or 8.
Example::
# your_trainer_file.py
# default used by the Trainer (ie: train on CPU)
trainer = Trainer(tpu_cores=None)
# int: train on a single core
trainer = Trainer(tpu_cores=1)
# list: train on a single selected core
trainer = Trainer(tpu_cores=[2])
# int: train on all cores few cores
trainer = Trainer(tpu_cores=8)
# for 8+ cores must submit via xla script with
# a max of 8 cores specified. The XLA script
# will duplicate script onto each TPU in the POD
trainer = Trainer(tpu_cores=8)
To train on more than 8 cores (ie: a POD),
submit this script using the xla_dist script.
Example::
python -m torch_xla.distributed.xla_dist
--tpu=$TPU_POD_NAME
--conda-env=torch-xla-nightly
--env=XLA_USE_BF16=1
-- python your_trainer_file.py
truncated_bptt_steps
^^^^^^^^^^^^^^^^^^^^
.. raw:: html
|
Truncated back prop breaks performs backprop every k steps of
a much longer sequence.
If this is enabled, your batches will automatically get truncated
and the trainer will apply Truncated Backprop to it.
(`Williams et al. "An efficient gradient-based algorithm for on-line training of
recurrent network trajectories."
`_)
.. testcode::
# default used by the Trainer (ie: disabled)
trainer = Trainer(truncated_bptt_steps=None)
# backprop every 5 steps in a batch
trainer = Trainer(truncated_bptt_steps=5)
.. note:: Make sure your batches have a sequence dimension.
Lightning takes care to split your batch along the time-dimension.
.. code-block:: python
# we use the second as the time dimension
# (batch, time, ...)
sub_batch = batch[0, 0:t, ...]
Using this feature requires updating your LightningModule's
:meth:`pytorch_lightning.core.LightningModule.training_step` to include a `hiddens` arg
with the hidden
.. code-block:: python
# Truncated back-propagation through time
def training_step(self, batch, batch_idx, hiddens):
# hiddens are the hiddens from the previous truncated backprop step
out, hiddens = self.lstm(data, hiddens)
# remember to detach() hiddens.
# If you don't, you will get a RuntimeError: Trying to backward through
# the graph a second time...
# Using hiddens.detach() allows each split to be disconnected.
return {
"loss": ...,
"hiddens": hiddens # remember to detach() this
}
To modify how the batch is split,
override :meth:`pytorch_lightning.core.LightningModule.tbptt_split_batch`:
.. testcode::
class LitMNIST(LightningModule):
def tbptt_split_batch(self, batch, split_size):
# do your own splitting on the batch
return splits
val_check_interval
^^^^^^^^^^^^^^^^^^
.. raw:: html
|
How often within one training epoch to check the validation set.
Can specify as float or int.
- use (float) to check within a training epoch
- use (int) to check every n steps (batches)
.. testcode::
# default used by the Trainer
trainer = Trainer(val_check_interval=1.0)
# check validation set 4 times during a training epoch
trainer = Trainer(val_check_interval=0.25)
# check validation set every 1000 training batches
# use this when using iterableDataset and your dataset has no length
# (ie: production cases with streaming data)
trainer = Trainer(val_check_interval=1000)
weights_save_path
^^^^^^^^^^^^^^^^^
.. raw:: html
|
Directory of where to save weights if specified.
.. testcode::
# default used by the Trainer
trainer = Trainer(weights_save_path=os.getcwd())
# save to your custom path
trainer = Trainer(weights_save_path='my/path')
Example::
# if checkpoint callback used, then overrides the weights path
# **NOTE: this saves weights to some/path NOT my/path
checkpoint = ModelCheckpoint(dirpath='some/path')
trainer = Trainer(
callbacks=[checkpoint],
weights_save_path='my/path'
)
weights_summary
^^^^^^^^^^^^^^^
.. raw:: html
|
Prints a summary of the weights when training begins.
Options: 'full', 'top', None.
.. testcode::
# default used by the Trainer (ie: print summary of top level modules)
trainer = Trainer(weights_summary='top')
# print full summary of all modules and submodules
trainer = Trainer(weights_summary='full')
# don't print a summary
trainer = Trainer(weights_summary=None)
-----
Trainer class API
-----------------
Methods
^^^^^^^
init
****
.. automethod:: pytorch_lightning.trainer.Trainer.__init__
:noindex:
fit
****
.. automethod:: pytorch_lightning.trainer.Trainer.fit
:noindex:
test
****
.. automethod:: pytorch_lightning.trainer.Trainer.test
:noindex:
tune
****
.. automethod:: pytorch_lightning.trainer.Trainer.tune
:noindex:
Properties
^^^^^^^^^^
callback_metrics
****************
The metrics available to callbacks. These are automatically set when you log via `self.log`
.. code-block:: python
def training_step(self, batch, batch_idx):
self.log('a_val', 2)
callback_metrics = trainer.callback_metrics
assert callback_metrics['a_val'] == 2
current_epoch
*************
The current epoch
.. code-block:: python
def training_step(self, batch, batch_idx):
current_epoch = self.trainer.current_epoch
if current_epoch > 100:
# do something
pass
logger (p)
**********
The current logger being used. Here's an example using tensorboard
.. code-block:: python
def training_step(self, batch, batch_idx):
logger = self.trainer.logger
tensorboard = logger.experiment
logged_metrics
**************
The metrics sent to the logger (visualizer).
.. code-block:: python
def training_step(self, batch, batch_idx):
self.log('a_val', 2, log=True)
logged_metrics = trainer.logged_metrics
assert logged_metrics['a_val'] == 2
log_dir
*******
The directory for the current experiment. Use this to save images to, etc...
.. code-block:: python
def training_step(self, batch, batch_idx):
img = ...
save_img(img, self.trainer.log_dir)
is_global_zero
**************
Whether this process is the global zero in multi-node training
.. code-block:: python
def training_step(self, batch, batch_idx):
if self.trainer.is_global_zero:
print('in node 0, accelerator 0')
progress_bar_metrics
********************
The metrics sent to the progress bar.
.. code-block:: python
def training_step(self, batch, batch_idx):
self.log('a_val', 2, prog_bar=True)
progress_bar_metrics = trainer.progress_bar_metrics
assert progress_bar_metrics['a_val'] == 2