if hparams["class_balanced"]:
in_weights = misc.make_weights_for_balanced_classes(in_)
out_weights = misc.make_weights_for_balanced_classes(out)
if uda is not None:
uda_weights = misc.make_weights_for_balanced_classes(uda)
else:
in_weights, out_weights, uda_weights = None, None, None
in_splits.append((in_, in_weights))
out_splits.append((out, out_weights))
if len(uda):
uda_splits.append((uda, uda_weights))
train_loaders = [
InfiniteDataLoader(
dataset=env,
weights=env_weights,
batch_size=hparams["batch_size"],
num_workers=dataset.N_WORKERS,
) for i, (env, env_weights) in enumerate(in_splits)
if i not in args.test_envs
]
uda_loaders = [
InfiniteDataLoader(
dataset=env,
weights=env_weights,
batch_size=hparams["batch_size"],
num_workers=dataset.N_WORKERS,
) for i, (env, env_weights) in enumerate(uda_splits)
if i in args.test_envs
]
if hparams['class_balanced']:
in_weights = misc.make_weights_for_balanced_classes(in_)
out_weights = misc.make_weights_for_balanced_classes(out)
if uda is not None:
uda_weights = misc.make_weights_for_balanced_classes(uda)
else:
in_weights, out_weights, uda_weights = None, None, None
in_splits.append((in_, in_weights))
out_splits.append((out, out_weights))
if len(uda):
uda_splits.append((uda, uda_weights))
train_loaders = [
InfiniteDataLoader(dataset=env,
weights=env_weights,
batch_size=hparams['batch_size'],
num_workers=dataset.N_WORKERS)
for i, (env, env_weights) in enumerate(in_splits)
if i not in args.test_envs
]
uda_loaders = [
InfiniteDataLoader(dataset=env,
weights=env_weights,
batch_size=hparams['batch_size'],
num_workers=dataset.N_WORKERS)
for i, (env, env_weights) in enumerate(uda_splits)
if i in args.test_envs
]
eval_loaders = [
out_splits = []
for env_i, env in enumerate(dataset):
out, in_ = misc.split_dataset(env,
int(len(env)*args.holdout_fraction),
misc.seed_hash(args.trial_seed, env_i))
if hparams['class_balanced']:
in_weights = misc.make_weights_for_balanced_classes(in_)
out_weights = misc.make_weights_for_balanced_classes(out)
else:
in_weights, out_weights = None, None
in_splits.append((in_, in_weights))
out_splits.append((out, out_weights))
train_loaders = [InfiniteDataLoader(
dataset=env,
weights=env_weights,
batch_size=hparams['batch_size'],
num_workers=dataset.N_WORKERS)
for i, (env, env_weights) in enumerate(in_splits)
if i not in args.test_envs]
# For UDA
test_in_loaders = [InfiniteDataLoader(
dataset=env,
weights=env_weights,
batch_size=hparams['batch_size'],
num_workers=dataset.N_WORKERS)
for i, (env, env_weights) in enumerate(in_splits)
if i in args.test_envs]
# For reporting