def test_ddp_sampler_error(tmpdir):
"""
Make sure DDP + AMP work
:return:
"""
if not tutils.can_run_gpu_test():
return
tutils.reset_seed()
tutils.set_random_master_port()
hparams = tutils.get_hparams()
model = LightningTestModel(hparams, force_remove_distributed_sampler=True)
logger = tutils.get_test_tube_logger(tmpdir, True)
trainer = Trainer(
logger=logger,
show_progress_bar=False,
max_nb_epochs=1,
gpus=[0, 1],
distributed_backend='ddp',
use_amp=True
)
with pytest.warns(UserWarning):
trainer.get_dataloaders(model)
def test_multi_gpu_model_dp(tmpdir):
"""
Make sure DP works
:return:
"""
tutils.reset_seed()
if not tutils.can_run_gpu_test():
return
model, hparams = tutils.get_model()
trainer_options = dict(
default_save_path=tmpdir,
show_progress_bar=False,
distributed_backend='dp',
max_nb_epochs=1,
train_percent_check=0.1,
val_percent_check=0.1,
gpus='-1'
)
tutils.run_model_test(trainer_options, model, hparams)
# test memory helper functions
memory.get_memory_profile('min_max')
def test_single_gpu_batch_parse():
tutils.reset_seed()
if not tutils.can_run_gpu_test():
return
trainer = Trainer()
# batch is just a tensor
batch = torch.rand(2, 3)
batch = trainer.transfer_batch_to_gpu(batch, 0)
assert batch.device.index == 0 and batch.type() == 'torch.cuda.FloatTensor'
# tensor list
batch = [torch.rand(2, 3), torch.rand(2, 3)]
batch = trainer.transfer_batch_to_gpu(batch, 0)
assert batch[0].device.index == 0 and batch[0].type(
) == 'torch.cuda.FloatTensor'
assert batch[1].device.index == 0 and batch[1].type(
) == 'torch.cuda.FloatTensor'
# tensor list of lists
batch = [[torch.rand(2, 3), torch.rand(2, 3)]]
batch = trainer.transfer_batch_to_gpu(batch, 0)
assert batch[0][0].device.index == 0 and batch[0][0].type(
) == 'torch.cuda.FloatTensor'
assert batch[0][1].device.index == 0 and batch[0][1].type(
) == 'torch.cuda.FloatTensor'
# tensor dict
batch = [{'a': torch.rand(2, 3), 'b': torch.rand(2, 3)}]
batch = trainer.transfer_batch_to_gpu(batch, 0)
assert batch[0]['a'].device.index == 0 and batch[0]['a'].type(
) == 'torch.cuda.FloatTensor'
assert batch[0]['b'].device.index == 0 and batch[0]['b'].type(
) == 'torch.cuda.FloatTensor'
# tuple of tensor list and list of tensor dict
batch = ([torch.rand(2, 3) for _ in range(2)], [{
'a': torch.rand(2, 3),
'b': torch.rand(2, 3)
} for _ in range(2)])
batch = trainer.transfer_batch_to_gpu(batch, 0)
assert batch[0][0].device.index == 0 and batch[0][0].type(
) == 'torch.cuda.FloatTensor'
assert batch[1][0]['a'].device.index == 0
assert batch[1][0]['a'].type() == 'torch.cuda.FloatTensor'
assert batch[1][0]['b'].device.index == 0
assert batch[1][0]['b'].type() == 'torch.cuda.FloatTensor'
def test_running_test_pretrained_model_ddp():
"""Verify test() on pretrained model"""
if not tutils.can_run_gpu_test():
return
tutils.reset_seed()
tutils.set_random_master_port()
hparams = tutils.get_hparams()
model = LightningTestModel(hparams)
save_dir = tutils.init_save_dir()
# exp file to get meta
logger = tutils.get_test_tube_logger(False)
# exp file to get weights
checkpoint = tutils.init_checkpoint_callback(logger)
trainer_options = dict(
show_progress_bar=False,
max_nb_epochs=1,
train_percent_check=0.4,
val_percent_check=0.2,
checkpoint_callback=checkpoint,
logger=logger,
gpus=[0, 1],
distributed_backend='ddp'
)
# fit model
trainer = Trainer(**trainer_options)
result = trainer.fit(model)
exp = logger.experiment
logging.info(os.listdir(exp.get_data_path(exp.name, exp.version)))
# correct result and ok accuracy
assert result == 1, 'training failed to complete'
pretrained_model = tutils.load_model(logger.experiment,
trainer.checkpoint_callback.filepath,
module_class=LightningTestModel)
# run test set
new_trainer = Trainer(**trainer_options)
new_trainer.test(pretrained_model)
for dataloader in model.test_dataloader():
tutils.run_prediction(dataloader, pretrained_model)
tutils.clear_save_dir()
def test_amp_gpu_dp(tmpdir):
"""Make sure DP + AMP work."""
tutils.reset_seed()
if not tutils.can_run_gpu_test():
return
model, hparams = tutils.get_model()
trainer_options = dict(
default_save_path=tmpdir,
max_num_epochs=1,
gpus='0, 1', # test init with gpu string
distributed_backend='dp',
use_amp=True)
with pytest.raises(MisconfigurationException):
tutils.run_model_test(trainer_options, model, hparams)
def test_multi_gpu_none_backend(tmpdir):
"""Make sure when using multiple GPUs the user can't use `distributed_backend = None`."""
tutils.reset_seed()
if not tutils.can_run_gpu_test():
return
model, hparams = tutils.get_model()
trainer_options = dict(default_save_path=tmpdir,
show_progress_bar=False,
max_num_epochs=1,
train_percent_check=0.1,
val_percent_check=0.1,
gpus='-1')
with pytest.raises(MisconfigurationException):
tutils.run_model_test(trainer_options, model)
def test_running_test_pretrained_model_dp():
tutils.reset_seed()
"""Verify test() on pretrained model"""
if not tutils.can_run_gpu_test():
return
hparams = tutils.get_hparams()
model = LightningTestModel(hparams)
save_dir = tutils.init_save_dir()
# logger file to get meta
logger = tutils.get_test_tube_logger(False)
# logger file to get weights
checkpoint = tutils.init_checkpoint_callback(logger)
trainer_options = dict(
show_progress_bar=True,
max_nb_epochs=1,
train_percent_check=0.4,
val_percent_check=0.2,
checkpoint_callback=checkpoint,
logger=logger,
gpus=[0, 1],
distributed_backend='dp'
)
# fit model
trainer = Trainer(**trainer_options)
result = trainer.fit(model)
# correct result and ok accuracy
assert result == 1, 'training failed to complete'
pretrained_model = tutils.load_model(logger.experiment,
trainer.checkpoint_callback.filepath,
module_class=LightningTestModel)
new_trainer = Trainer(**trainer_options)
new_trainer.test(pretrained_model)
# test we have good test accuracy
tutils.assert_ok_test_acc(new_trainer)
tutils.clear_save_dir()
def test_amp_single_gpu(tmpdir):
"""Make sure DDP + AMP work."""
tutils.reset_seed()
if not tutils.can_run_gpu_test():
return
hparams = tutils.get_hparams()
model = LightningTestModel(hparams)
trainer_options = dict(default_save_path=tmpdir,
show_progress_bar=True,
max_epochs=1,
gpus=1,
distributed_backend='ddp',
use_amp=True)
tutils.run_model_test(trainer_options, model)
def test_multi_gpu_model_ddp(tmpdir):
"""Make sure DDP works."""
if not tutils.can_run_gpu_test():
return
tutils.reset_seed()
tutils.set_random_master_port()
model, hparams = tutils.get_model()
trainer_options = dict(default_save_path=tmpdir,
show_progress_bar=False,
max_num_epochs=1,
train_percent_check=0.4,
val_percent_check=0.2,
gpus=[0, 1],
distributed_backend='ddp')
tutils.run_model_test(trainer_options, model)
def test_amp_gpu_ddp_slurm_managed(tmpdir):
"""Make sure DDP + AMP work."""
if not tutils.can_run_gpu_test():
return
tutils.reset_seed()
# simulate setting slurm flags
tutils.set_random_master_port()
os.environ['SLURM_LOCALID'] = str(0)
hparams = tutils.get_hparams()
model = LightningTestModel(hparams)
trainer_options = dict(show_progress_bar=True,
max_epochs=1,
gpus=[0],
distributed_backend='ddp',
use_amp=True)
# exp file to get meta
logger = tutils.get_test_tube_logger(tmpdir, False)
# exp file to get weights
checkpoint = tutils.init_checkpoint_callback(logger)
# add these to the trainer options
trainer_options['checkpoint_callback'] = checkpoint
trainer_options['logger'] = logger
# fit model
trainer = Trainer(**trainer_options)
trainer.is_slurm_managing_tasks = True
result = trainer.fit(model)
# correct result and ok accuracy
assert result == 1, 'amp + ddp model failed to complete'
# test root model address
assert trainer.resolve_root_node_address('abc') == 'abc'
assert trainer.resolve_root_node_address('abc[23]') == 'abc23'
assert trainer.resolve_root_node_address('abc[23-24]') == 'abc23'
assert trainer.resolve_root_node_address(
'abc[23-24, 45-40, 40]') == 'abc23'
def test_amp_gpu_dp(tmpdir):
"""Make sure DP + AMP work."""
tutils.reset_seed()
if not tutils.can_run_gpu_test():
return
model, hparams = tutils.get_model()
trainer_options = dict(
default_save_path=tmpdir,
max_epochs=1,
gpus='0, 1', # test init with gpu string
distributed_backend='dp',
use_amp=True)
trainer = Trainer(**trainer_options)
result = trainer.fit(model)
assert result == 1
def test_no_amp_single_gpu(tmpdir):
"""Make sure DDP + AMP work."""
tutils.reset_seed()
if not tutils.can_run_gpu_test():
return
hparams = tutils.get_hparams()
model = LightningTestModel(hparams)
trainer_options = dict(default_save_path=tmpdir,
show_progress_bar=True,
max_num_epochs=1,
gpus=1,
distributed_backend='dp',
use_amp=True)
with pytest.raises((MisconfigurationException, ModuleNotFoundError)):
tutils.run_model_test(trainer_options, model)
def test_amp_gpu_ddp():
"""
Make sure DDP + AMP work
:return:
"""
if not tutils.can_run_gpu_test():
return
tutils.reset_seed()
tutils.set_random_master_port()
hparams = tutils.get_hparams()
model = LightningTestModel(hparams)
trainer_options = dict(show_progress_bar=True,
max_nb_epochs=1,
gpus=2,
distributed_backend='ddp',
use_amp=True)
tutils.run_model_test(trainer_options, model, hparams)
def test_multi_gpu_model_ddp2():
"""
Make sure DDP2 works
:return:
"""
if not tutils.can_run_gpu_test():
return
tutils.reset_seed()
tutils.set_random_master_port()
model, hparams = tutils.get_model()
trainer_options = dict(show_progress_bar=True,
max_nb_epochs=1,
train_percent_check=0.4,
val_percent_check=0.2,
gpus=2,
weights_summary=None,
distributed_backend='ddp2')
tutils.run_model_test(trainer_options, model, hparams)
def test_dp_resume():
"""
Make sure DP continues training correctly
:return:
"""
if not tutils.can_run_gpu_test():
return
tutils.reset_seed()
hparams = tutils.get_hparams()
model = LightningTestModel(hparams)
trainer_options = dict(
show_progress_bar=True,
max_nb_epochs=2,
gpus=2,
distributed_backend='dp',
)
save_dir = tutils.init_save_dir()
# get logger
logger = tutils.get_test_tube_logger(debug=False)
# exp file to get weights
# logger file to get weights
checkpoint = tutils.init_checkpoint_callback(logger)
# add these to the trainer options
trainer_options['logger'] = logger
trainer_options['checkpoint_callback'] = checkpoint
# fit model
trainer = Trainer(**trainer_options)
trainer.is_slurm_managing_tasks = True
result = trainer.fit(model)
# track epoch before saving
real_global_epoch = trainer.current_epoch
# correct result and ok accuracy
assert result == 1, 'amp + dp model failed to complete'
# ---------------------------
# HPC LOAD/SAVE
# ---------------------------
# save
trainer.hpc_save(save_dir, logger)
# init new trainer
new_logger = tutils.get_test_tube_logger(version=logger.version)
trainer_options['logger'] = new_logger
trainer_options['checkpoint_callback'] = ModelCheckpoint(save_dir)
trainer_options['train_percent_check'] = 0.2
trainer_options['val_percent_check'] = 0.2
trainer_options['max_nb_epochs'] = 1
new_trainer = Trainer(**trainer_options)
# set the epoch start hook so we can predict before the model does the full training
def assert_good_acc():
assert new_trainer.current_epoch == real_global_epoch and new_trainer.current_epoch > 0
# if model and state loaded correctly, predictions will be good even though we
# haven't trained with the new loaded model
dp_model = new_trainer.model
dp_model.eval()
dataloader = trainer.get_train_dataloader()
tutils.run_prediction(dataloader, dp_model, dp=True)
# new model
model = LightningTestModel(hparams)
model.on_sanity_check_start = assert_good_acc
# fit new model which should load hpc weights
new_trainer.fit(model)
# test freeze on gpu
model.freeze()
model.unfreeze()
tutils.clear_save_dir()
def test_amp_gpu_ddp_slurm_managed():
"""
Make sure DDP + AMP work
:return:
"""
if not tutils.can_run_gpu_test():
return
tutils.reset_seed()
# simulate setting slurm flags
tutils.set_random_master_port()
os.environ['SLURM_LOCALID'] = str(0)
hparams = tutils.get_hparams()
model = LightningTestModel(hparams)
trainer_options = dict(show_progress_bar=True,
max_nb_epochs=1,
gpus=[0],
distributed_backend='ddp',
use_amp=True)
save_dir = tutils.init_save_dir()
# exp file to get meta
logger = tutils.get_test_tube_logger(False)
# exp file to get weights
checkpoint = tutils.init_checkpoint_callback(logger)
# add these to the trainer options
trainer_options['checkpoint_callback'] = checkpoint
trainer_options['logger'] = logger
# fit model
trainer = Trainer(**trainer_options)
trainer.is_slurm_managing_tasks = True
result = trainer.fit(model)
# correct result and ok accuracy
assert result == 1, 'amp + ddp model failed to complete'
# test root model address
assert trainer.resolve_root_node_address('abc') == 'abc'
assert trainer.resolve_root_node_address('abc[23]') == 'abc23'
assert trainer.resolve_root_node_address('abc[23-24]') == 'abc23'
assert trainer.resolve_root_node_address(
'abc[23-24, 45-40, 40]') == 'abc23'
# test model loading with a map_location
pretrained_model = tutils.load_model(logger.experiment,
trainer.checkpoint_callback.filepath)
# test model preds
for dataloader in trainer.get_test_dataloaders():
tutils.run_prediction(dataloader, pretrained_model)
if trainer.use_ddp:
# on hpc this would work fine... but need to hack it for the purpose of the test
trainer.model = pretrained_model
trainer.optimizers, trainer.lr_schedulers = pretrained_model.configure_optimizers(
)
# test HPC loading / saving
trainer.hpc_save(save_dir, logger)
trainer.hpc_load(save_dir, on_gpu=True)
# test freeze on gpu
model.freeze()
model.unfreeze()
tutils.clear_save_dir()
