roidb = merge_roidb(roidbs)
roidb = filter_roidb(roidb, config)
bbox_means, bbox_stds = add_bbox_regression_targets(roidb, config)
print('Creating Iterator with {} Images'.format(len(roidb)))
#horovod DataIter
#train_iter=[]
#for i in range(num_workers):
worker_roidb = random.sample(roidb, len(roidb) // num_workers)
train_iter = MNIteratorE2E(roidb=worker_roidb,
config=config,
batch_size=batch_size,
nGPUs=nGPUs,
threads=config.TRAIN.NUM_THREAD,
pad_rois_to=400)
# train_iter.append(worker_train_iter)
print('The Iterator has {} samples!'.format(len(train_iter)))
# Creating the Logger
if rank == 0:
logger, output_path = create_logger(config.output_path, args.cfg,
config.dataset.image_set)
# get list of fixed parameters
print('Initializing the model...')
sym_inst = eval('{}.{}'.format(config.symbol,
image_sets = [iset for iset in config.dataset.image_set.split('+')]
roidbs = [load_proposal_roidb(config.dataset.dataset, image_set, config.dataset.root_path,
config.dataset.dataset_path,
proposal=config.dataset.proposal, append_gt=True, flip=config.TRAIN.FLIP,
result_path=config.output_path,
proposal_path=config.proposal_path, load_mask=config.TRAIN.WITH_MASK, only_gt=not config.TRAIN.USE_NEG_CHIPS)
for image_set in image_sets]
roidb = merge_roidb(roidbs)
roidb = filter_roidb(roidb, config)
bbox_means, bbox_stds = add_bbox_regression_targets(roidb, config)
print('Creating Iterator with {} Images'.format(len(roidb)))
train_iter = MNIteratorE2E(roidb=roidb, config=config, batch_size=batch_size, nGPUs=nGPUs,
threads=config.TRAIN.NUM_THREAD, pad_rois_to=400) #, crop_size=(config.TRAIN.SCALES[-1],config.TRAIN.SCALES[-1]))
print('The Iterator has {} samples!'.format(len(train_iter)))
# Creating the Logger
logger, output_path = create_logger(config.output_path, args.cfg, config.dataset.image_set)
# get list of fixed parameters
print('Initializing the model...')
sym_inst = eval('{}.{}'.format(config.symbol, config.symbol))(n_proposals=400, momentum=args.momentum)
sym = sym_inst.get_symbol_rcnn(config)
fixed_param_names = get_fixed_param_names(config.network.FIXED_PARAMS, sym)
# Creating the module
mod = mx.mod.Module(symbol=sym,
context=context,
result_path=config.output_path,
proposal_path=config.proposal_path,
load_mask=config.TRAIN.WITH_MASK,
only_gt=not config.TRAIN.USE_NEG_CHIPS)
for image_set in image_sets
]
roidb = merge_roidb(roidbs)
roidb = filter_roidb(roidb, config)
bbox_means, bbox_stds = add_bbox_regression_targets(roidb, config)
print('Creating Iterator with {} Images'.format(len(roidb)))
train_iter = MNIteratorE2E(roidb=roidb,
config=config,
batch_size=batch_size,
nGPUs=nGPUs,
threads=config.TRAIN.NUM_THREAD,
pad_rois_to=400,
crop_size=(216, 216))
print('The Iterator has {} samples!'.format(len(train_iter)))
# Creating the Logger
logger, output_path = create_logger(config.output_path, args.cfg,
config.dataset.image_set)
# get list of fixed parameters
print('Initializing the model...')
sym_inst = eval('{}.{}'.format(config.symbol,
config.symbol))(n_proposals=400,
momentum=args.momentum)
sym = sym_inst.get_symbol_rpn_ugly(config)
