Closed anshkumar closed 3 years ago
The problem was with the data pipeline. When I changed it to the following, it worked:
img_norm_cfg = dict(mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='LoadAnnotations', with_bbox=True),
dict(type='Resize', img_scale=(750, 750), keep_ratio=True),
dict(type='RandomFlip', flip_ratio=0.5),
dict(type='Normalize', **img_norm_cfg),
dict(type='DefaultFormatBundle'),
dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels'])
]
test_pipeline = [
dict(type='LoadImageFromFile'),
dict(
type='MultiScaleFlipAug',
img_scale=(750, 750),
flip=False,
transforms=[
dict(type='Resize', keep_ratio=True),
dict(type='RandomFlip'),
dict(type='Normalize',**img_norm_cfg),
dict(type='Pad', size_divisor=32),
dict(type='ImageToTensor', keys=['img']),
dict(type='Collect', keys=['img'])
])
]
@ZwwWayne @anshkumar I have the same problem and I am unable to solve. despite specifying my datasets and inheriting from base, I am lost as to how I should proceed and how configs are to be created. :( they seem pretty complicated...
I'm getting following error while trying to train a model.
My training config is as follows:
I'm using a custom dataset.
Environment
Error traceback
My custom training json is this.