Image-to-Text
HTRflow
Swedish
satrn_htr / config.py
Gabriel's picture
Update config.py
dd23264
raw
history blame
13.8 kB
default_scope = 'mmocr'
env_cfg = dict(
cudnn_benchmark=True,
mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0),
dist_cfg=dict(backend='nccl'))
randomness = dict(seed=None)
default_hooks = dict(
timer=dict(type='IterTimerHook'),
logger=dict(type='LoggerHook', interval=100),
param_scheduler=dict(type='ParamSchedulerHook'),
checkpoint=dict(type='CheckpointHook', interval=1),
sampler_seed=dict(type='DistSamplerSeedHook'),
sync_buffer=dict(type='SyncBuffersHook'),
visualization=dict(
type='VisualizationHook',
interval=1,
enable=False,
show=False,
draw_gt=False,
draw_pred=False))
log_level = 'INFO'
log_processor = dict(type='LogProcessor', window_size=10, by_epoch=True)
load_from = './model.pth'
resume = False
val_evaluator = dict(
type='Evaluator',
metrics=[
dict(
type='WordMetric',
mode=['exact', 'ignore_case', 'ignore_case_symbol'],
valid_symbol='[^A-Z^a-z^0-9^一-龥^å^ä^ö^Å^Ä^Ö]'),
dict(type='CharMetric', valid_symbol='[^A-Z^a-z^0-9^一-龥^å^ä^ö^Å^Ä^Ö]'),
dict(
type='OneMinusNEDMetric',
valid_symbol='[^A-Z^a-z^0-9^一-龥^å^ä^ö^Å^Ä^Ö]')
])
test_evaluator = dict(
type='Evaluator',
metrics=[
dict(
type='WordMetric',
mode=['exact', 'ignore_case', 'ignore_case_symbol'],
valid_symbol='[^A-Z^a-z^0-9^一-龥^å^ä^ö^Å^Ä^Ö]'),
dict(type='CharMetric', valid_symbol='[^A-Z^a-z^0-9^一-龥^å^ä^ö^Å^Ä^Ö]'),
dict(
type='OneMinusNEDMetric',
valid_symbol='[^A-Z^a-z^0-9^一-龥^å^ä^ö^Å^Ä^Ö]')
])
vis_backends = [dict(type='LocalVisBackend')]
visualizer = dict(
type='TextRecogLocalVisualizer',
name='visualizer',
vis_backends=[dict(type='TensorboardVisBackend')])
optim_wrapper = dict(
type='OptimWrapper', optimizer=dict(type='Adam', lr=0.0003))
train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=5, val_interval=1)
val_cfg = dict(type='ValLoop')
test_cfg = dict(type='TestLoop')
param_scheduler = [dict(type='MultiStepLR', milestones=[3, 4], end=5)]
file_client_args = dict(backend='disk')
dictionary = dict(
type='Dictionary',
dict_file=
'./models--Riksarkivet--HTR_pipeline_models/snapshots/296681baf68583f07e89b5fed08136b77e3904cd/SATRN/dict1700.txt',
with_padding=True,
with_unknown=True,
same_start_end=True,
with_start=True,
with_end=True)
model = dict(
type='SATRN',
backbone=dict(type='ShallowCNN', input_channels=3, hidden_dim=512),
encoder=dict(
type='SATRNEncoder',
n_layers=12,
n_head=8,
d_k=64,
d_v=64,
d_model=512,
n_position=100,
d_inner=2048,
dropout=0.1),
decoder=dict(
type='NRTRDecoder',
n_layers=6,
d_embedding=512,
n_head=8,
d_model=512,
d_inner=2048,
d_k=64,
d_v=64,
module_loss=dict(
type='CEModuleLoss', flatten=True, ignore_first_char=True),
dictionary=dict(
type='Dictionary',
dict_file=
'./models--Riksarkivet--HTR_pipeline_models/snapshots/296681baf68583f07e89b5fed08136b77e3904cd/SATRN/dict1700.txt',
with_padding=True,
with_unknown=True,
same_start_end=True,
with_start=True,
with_end=True),
max_seq_len=100,
postprocessor=dict(type='AttentionPostprocessor')),
data_preprocessor=dict(
type='TextRecogDataPreprocessor',
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375]))
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=dict(backend='disk'),
ignore_empty=True,
min_size=2),
dict(type='LoadOCRAnnotations', with_text=True),
dict(type='Resize', scale=(400, 64), keep_ratio=False),
dict(
type='PackTextRecogInputs',
meta_keys=('img_path', 'ori_shape', 'img_shape', 'valid_ratio'))
]
test_pipeline = [
dict(type='LoadImageFromFile', file_client_args=dict(backend='disk')),
dict(type='Resize', scale=(400, 64), keep_ratio=False),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
type='PackTextRecogInputs',
meta_keys=('img_path', 'ori_shape', 'img_shape', 'valid_ratio'))
]
HTR_1700_combined_train = dict(
type='RecogTextDataset',
parser_cfg=dict(type='LineJsonParser', keys=['filename', 'text']),
data_root='/ceph/hpc/scratch/user/euerikl/data/HTR_1700_clean',
ann_file=
'/ceph/hpc/home/euerikl/projects/hf_openmmlab_models/data/processed/1700_HTR_shuffled_train.jsonl',
test_mode=False,
pipeline=None)
HTR_1700_combined_test = dict(
type='RecogTextDataset',
parser_cfg=dict(type='LineJsonParser', keys=['filename', 'text']),
data_root='/ceph/hpc/scratch/user/euerikl/data/HTR_1700_clean',
ann_file=
'/ceph/hpc/home/euerikl/projects/hf_openmmlab_models/data/processed/1700_HTR_shuffled_val.jsonl',
test_mode=True,
pipeline=None)
pr_cr_combined_train = dict(
type='RecogTextDataset',
parser_cfg=dict(
type='LineStrParser', keys=['filename', 'text'], separator='|'),
data_root='/ceph/hpc/scratch/user/euerikl/data/line_images',
ann_file=
'/ceph/hpc/home/euerikl/projects/htr_1800/gt_files/combined_train.txt',
test_mode=False,
pipeline=None)
pr_cr_combined_test = dict(
type='RecogTextDataset',
parser_cfg=dict(
type='LineStrParser', keys=['filename', 'text'], separator='|'),
data_root='/ceph/hpc/scratch/user/euerikl/data/line_images',
ann_file=
'/ceph/hpc/home/euerikl/projects/htr_1800/gt_files/combined_eval.txt',
test_mode=True,
pipeline=None)
out_of_domain_1700_all_test = dict(
type='RecogTextDataset',
parser_cfg=dict(type='LineJsonParser', keys=['filename', 'text']),
data_root='/ceph/hpc/scratch/user/euerikl/data/HTR_1700_testsets_clean',
ann_file=
'/ceph/hpc/home/euerikl/projects/hf_openmmlab_models/data/processed/1700_testsets_gt/1700_HTR_testsets_all.jsonl',
test_mode=True,
pipeline=None)
train_list = [
dict(
type='RecogTextDataset',
parser_cfg=dict(type='LineJsonParser', keys=['filename', 'text']),
data_root='/ceph/hpc/scratch/user/euerikl/data/HTR_1700_clean',
ann_file=
'/ceph/hpc/home/euerikl/projects/hf_openmmlab_models/data/processed/1700_HTR_shuffled_train.jsonl',
test_mode=False,
pipeline=None),
dict(
type='RecogTextDataset',
parser_cfg=dict(
type='LineStrParser', keys=['filename', 'text'], separator='|'),
data_root='/ceph/hpc/scratch/user/euerikl/data/line_images',
ann_file=
'/ceph/hpc/home/euerikl/projects/htr_1800/gt_files/combined_train.txt',
test_mode=False,
pipeline=None)
]
test_list = [
dict(
type='RecogTextDataset',
parser_cfg=dict(type='LineJsonParser', keys=['filename', 'text']),
data_root='/ceph/hpc/scratch/user/euerikl/data/HTR_1700_testsets_clean',
ann_file=
'/ceph/hpc/home/euerikl/projects/hf_openmmlab_models/data/processed/1700_testsets_gt/1700_HTR_testsets_all.jsonl',
test_mode=True,
pipeline=None)
]
train_dataset = dict(
type='ConcatDataset',
datasets=[
dict(
type='RecogTextDataset',
parser_cfg=dict(type='LineJsonParser', keys=['filename', 'text']),
data_root='/ceph/hpc/scratch/user/euerikl/data/HTR_1700_clean',
ann_file=
'/ceph/hpc/home/euerikl/projects/hf_openmmlab_models/data/processed/1700_HTR_shuffled_train.jsonl',
test_mode=False,
pipeline=None),
dict(
type='RecogTextDataset',
parser_cfg=dict(
type='LineStrParser', keys=['filename', 'text'],
separator='|'),
data_root='/ceph/hpc/scratch/user/euerikl/data/line_images',
ann_file=
'/ceph/hpc/home/euerikl/projects/htr_1800/gt_files/combined_train.txt',
test_mode=False,
pipeline=None)
],
pipeline=[
dict(
type='LoadImageFromFile',
file_client_args=dict(backend='disk'),
ignore_empty=True,
min_size=2),
dict(type='LoadOCRAnnotations', with_text=True),
dict(type='Resize', scale=(400, 64), keep_ratio=False),
dict(
type='PackTextRecogInputs',
meta_keys=('img_path', 'ori_shape', 'img_shape', 'valid_ratio'))
])
test_dataset = dict(
type='ConcatDataset',
datasets=[
dict(
type='RecogTextDataset',
parser_cfg=dict(type='LineJsonParser', keys=['filename', 'text']),
data_root=
'/ceph/hpc/scratch/user/euerikl/data/HTR_1700_testsets_clean',
ann_file=
'/ceph/hpc/home/euerikl/projects/hf_openmmlab_models/data/processed/1700_testsets_gt/1700_HTR_testsets_all.jsonl',
test_mode=True,
pipeline=None)
],
pipeline=[
dict(type='LoadImageFromFile', file_client_args=dict(backend='disk')),
dict(type='Resize', scale=(400, 64), keep_ratio=False),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
type='PackTextRecogInputs',
meta_keys=('img_path', 'ori_shape', 'img_shape', 'valid_ratio'))
])
train_dataloader = dict(
batch_size=8,
num_workers=1,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=True),
dataset=dict(
type='ConcatDataset',
datasets=[
dict(
type='RecogTextDataset',
parser_cfg=dict(
type='LineJsonParser', keys=['filename', 'text']),
data_root='/ceph/hpc/scratch/user/euerikl/data/HTR_1700_clean',
ann_file=
'/ceph/hpc/home/euerikl/projects/hf_openmmlab_models/data/processed/1700_HTR_shuffled_train.jsonl',
test_mode=False,
pipeline=None),
dict(
type='RecogTextDataset',
parser_cfg=dict(
type='LineStrParser',
keys=['filename', 'text'],
separator='|'),
data_root='/ceph/hpc/scratch/user/euerikl/data/line_images',
ann_file=
'/ceph/hpc/home/euerikl/projects/htr_1800/gt_files/combined_train.txt',
test_mode=False,
pipeline=None)
],
pipeline=[
dict(
type='LoadImageFromFile',
file_client_args=dict(backend='disk'),
ignore_empty=True,
min_size=2),
dict(type='LoadOCRAnnotations', with_text=True),
dict(type='Resize', scale=(400, 64), keep_ratio=False),
dict(
type='PackTextRecogInputs',
meta_keys=('img_path', 'ori_shape', 'img_shape',
'valid_ratio'))
]))
test_dataloader = dict(
batch_size=8,
num_workers=1,
persistent_workers=True,
drop_last=False,
sampler=dict(type='DefaultSampler', shuffle=False),
dataset=dict(
type='ConcatDataset',
datasets=[
dict(
type='RecogTextDataset',
parser_cfg=dict(
type='LineJsonParser', keys=['filename', 'text']),
data_root=
'/ceph/hpc/scratch/user/euerikl/data/HTR_1700_testsets_clean',
ann_file=
'/ceph/hpc/home/euerikl/projects/hf_openmmlab_models/data/processed/1700_testsets_gt/1700_HTR_testsets_all.jsonl',
test_mode=True,
pipeline=None)
],
pipeline=[
dict(
type='LoadImageFromFile',
file_client_args=dict(backend='disk')),
dict(type='Resize', scale=(400, 64), keep_ratio=False),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
type='PackTextRecogInputs',
meta_keys=('img_path', 'ori_shape', 'img_shape',
'valid_ratio'))
]))
val_dataloader = dict(
batch_size=8,
num_workers=1,
persistent_workers=True,
drop_last=False,
sampler=dict(type='DefaultSampler', shuffle=False),
dataset=dict(
type='ConcatDataset',
datasets=[
dict(
type='RecogTextDataset',
parser_cfg=dict(
type='LineJsonParser', keys=['filename', 'text']),
data_root=
'/ceph/hpc/scratch/user/euerikl/data/HTR_1700_testsets_clean',
ann_file=
'/ceph/hpc/home/euerikl/projects/hf_openmmlab_models/data/processed/1700_testsets_gt/1700_HTR_testsets_all.jsonl',
test_mode=True,
pipeline=None)
],
pipeline=[
dict(
type='LoadImageFromFile',
file_client_args=dict(backend='disk')),
dict(type='Resize', scale=(400, 64), keep_ratio=False),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
type='PackTextRecogInputs',
meta_keys=('img_path', 'ori_shape', 'img_shape',
'valid_ratio'))
]))
gpu_ids = range(0, 4)
cudnn_benchmark = True
work_dir = '/ceph/hpc/home/euerikl/projects/hf_openmmlab_models/models/checkpoints/1700_1800_combined_satrn'
checkpoint_config = dict(interval=1)
auto_scale_lr = dict(base_batch_size=32)
launcher = 'pytorch'