def test_train_step(self):
config = detr_cfg.DetectionConfig(num_encoder_layers=1,
num_decoder_layers=1,
train_data=coco.COCODataConfig(
tfds_name='coco/2017',
tfds_split='validation',
is_training=True,
global_batch_size=2,
))
with tfds.testing.mock_data(as_dataset_fn=_as_dataset):
task = detection.DectectionTask(config)
model = task.build_model()
dataset = task.build_inputs(config.train_data)
iterator = iter(dataset)
opt_cfg = optimization.OptimizationConfig({
'optimizer': {
'type': 'detr_adamw',
'detr_adamw': {
'weight_decay_rate': 1e-4,
'global_clipnorm': 0.1,
}
},
'learning_rate': {
'type': 'stepwise',
'stepwise': {
'boundaries': [120000],
'values': [0.0001, 1.0e-05]
}
},
})
optimizer = detection.DectectionTask.create_optimizer(opt_cfg)
task.train_step(next(iterator), model, optimizer)
def detr_coco() -> cfg.ExperimentConfig:
"""Config to get results that matches the paper."""
train_batch_size = 64
eval_batch_size = 64
num_train_data = COCO_TRAIN_EXAMPLES
num_steps_per_epoch = num_train_data // train_batch_size
train_steps = 500 * num_steps_per_epoch # 500 epochs
decay_at = train_steps - 100 * num_steps_per_epoch # 400 epochs
config = cfg.ExperimentConfig(
task=DetrTask(init_checkpoint='',
init_checkpoint_modules='backbone',
model=Detr(num_classes=81,
input_size=[1333, 1333, 3],
norm_activation=common.NormActivation()),
losses=Losses(),
train_data=coco.COCODataConfig(
tfds_name='coco/2017',
tfds_split='train',
is_training=True,
global_batch_size=train_batch_size,
shuffle_buffer_size=1000,
),
validation_data=coco.COCODataConfig(
tfds_name='coco/2017',
tfds_split='validation',
is_training=False,
global_batch_size=eval_batch_size,
drop_remainder=False)),
trainer=cfg.TrainerConfig(
train_steps=train_steps,
validation_steps=-1,
steps_per_loop=10000,
summary_interval=10000,
checkpoint_interval=10000,
validation_interval=10000,
max_to_keep=1,
best_checkpoint_export_subdir='best_ckpt',
best_checkpoint_eval_metric='AP',
optimizer_config=optimization.OptimizationConfig({
'optimizer': {
'type': 'detr_adamw',
'detr_adamw': {
'weight_decay_rate': 1e-4,
'global_clipnorm': 0.1,
# Avoid AdamW legacy behavior.
'gradient_clip_norm': 0.0
}
},
'learning_rate': {
'type': 'stepwise',
'stepwise': {
'boundaries': [decay_at],
'values': [0.0001, 1.0e-05]
}
},
})),
restrictions=[
'task.train_data.is_training != None',
])
return config
def test_train_step(self):
config = detr_cfg.DetrTask(
model=detr_cfg.Detr(
input_size=[1333, 1333, 3],
num_encoder_layers=1,
num_decoder_layers=1,
backbone=backbones.Backbone(
type='resnet',
resnet=backbones.ResNet(model_id=10, bn_trainable=False))
),
losses=detr_cfg.Losses(class_offset=1),
train_data=detr_cfg.DataConfig(
tfds_name='coco/2017',
tfds_split='validation',
is_training=True,
global_batch_size=2,
))
with tfds.testing.mock_data(as_dataset_fn=_as_dataset):
task = detection.DetectionTask(config)
model = task.build_model()
dataset = task.build_inputs(config.train_data)
iterator = iter(dataset)
opt_cfg = optimization.OptimizationConfig({
'optimizer': {
'type': 'detr_adamw',
'detr_adamw': {
'weight_decay_rate': 1e-4,
'global_clipnorm': 0.1,
}
},
'learning_rate': {
'type': 'stepwise',
'stepwise': {
'boundaries': [120000],
'values': [0.0001, 1.0e-05]
}
},
})
optimizer = detection.DetectionTask.create_optimizer(opt_cfg)
task.train_step(next(iterator), model, optimizer)
def detr_coco_tfrecord() -> cfg.ExperimentConfig:
"""Config to get results that matches the paper."""
train_batch_size = 64
eval_batch_size = 64
steps_per_epoch = COCO_TRAIN_EXAMPLES // train_batch_size
train_steps = 300 * steps_per_epoch # 300 epochs
decay_at = train_steps - 100 * steps_per_epoch # 200 epochs
config = cfg.ExperimentConfig(
task=DetrTask(init_checkpoint='',
init_checkpoint_modules='backbone',
annotation_file=os.path.join(COCO_INPUT_PATH_BASE,
'instances_val2017.json'),
model=Detr(input_size=[1333, 1333, 3],
norm_activation=common.NormActivation()),
losses=Losses(),
train_data=DataConfig(
input_path=os.path.join(COCO_INPUT_PATH_BASE,
'train*'),
is_training=True,
global_batch_size=train_batch_size,
shuffle_buffer_size=1000,
),
validation_data=DataConfig(
input_path=os.path.join(COCO_INPUT_PATH_BASE,
'val*'),
is_training=False,
global_batch_size=eval_batch_size,
drop_remainder=False,
)),
trainer=cfg.TrainerConfig(
train_steps=train_steps,
validation_steps=COCO_VAL_EXAMPLES // eval_batch_size,
steps_per_loop=steps_per_epoch,
summary_interval=steps_per_epoch,
checkpoint_interval=steps_per_epoch,
validation_interval=5 * steps_per_epoch,
max_to_keep=1,
best_checkpoint_export_subdir='best_ckpt',
best_checkpoint_eval_metric='AP',
optimizer_config=optimization.OptimizationConfig({
'optimizer': {
'type': 'detr_adamw',
'detr_adamw': {
'weight_decay_rate': 1e-4,
'global_clipnorm': 0.1,
# Avoid AdamW legacy behavior.
'gradient_clip_norm': 0.0
}
},
'learning_rate': {
'type': 'stepwise',
'stepwise': {
'boundaries': [decay_at],
'values': [0.0001, 1.0e-05]
}
},
})),
restrictions=[
'task.train_data.is_training != None',
])
return config