mmpretrain/configs/_base_/datasets/coco_okvqa.py

76 lines
1.9 KiB
Python

# data settings
data_preprocessor = dict(
mean=[122.770938, 116.7460125, 104.09373615],
std=[68.5005327, 66.6321579, 70.32316305],
to_rgb=True,
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(
type='RandomResizedCrop',
scale=384,
interpolation='bicubic',
backend='pillow'),
dict(
type='PackInputs',
algorithm_keys=['question', 'gt_answer', 'gt_answer_weight'],
meta_keys=['question_id', 'image_id'],
),
]
test_pipeline = [
dict(type='LoadImageFromFile'),
dict(
type='Resize',
scale=(480, 480),
interpolation='bicubic',
backend='pillow'),
dict(
type='CleanCaption',
keys=['question'],
),
dict(
type='PackInputs',
algorithm_keys=['question', 'gt_answer', 'gt_answer_weight'],
meta_keys=['question_id', 'image_id'],
),
]
train_dataloader = dict(
batch_size=16,
num_workers=8,
dataset=dict(
type='COCOVQA',
data_root='data/coco',
data_prefix='train2014',
question_file=
'annotations/okvqa_OpenEnded_mscoco_train2014_questions.json',
ann_file='annotations/okvqa_mscoco_train2014_annotations.json',
pipeline=train_pipeline),
sampler=dict(type='DefaultSampler', shuffle=True),
persistent_workers=True,
drop_last=True,
)
val_dataloader = dict(
batch_size=16,
num_workers=8,
dataset=dict(
type='COCOVQA',
data_root='data/coco',
data_prefix='val2014',
question_file=
'annotations/okvqa_OpenEnded_mscoco_val2014_questions.json',
ann_file='annotations/okvqa_mscoco_val2014_annotations.json',
pipeline=test_pipeline),
sampler=dict(type='DefaultSampler', shuffle=False),
persistent_workers=True,
)
val_evaluator = dict(type='VQAAcc')
test_dataloader = val_dataloader
test_evaluator = val_evaluator