mirrors/mmsegmentation
1
0
Fork 0
mirror of https://github.com/open-mmlab/mmsegmentation.git synced 2025-06-03 22:03:48 +08:00
Code Issues Projects Releases Wiki Activity

[Fix] Fix batch size of val_dataloader to 1.

Browse Source
This commit is contained in:
limengzhang.vendor 2022-06-12 09:10:26 +00:00 committed by zhengmiao
parent a0fe318451
commit 34711422ac
67 changed files with 68 additions and 68 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
configs/_base_/datasets/ade20k.py
View File

@ -28,7 +28,7 @@ train_dataloader = dict(
img_path='images/training', seg_map_path='annotations/training'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/ade20k_640x640.py
View File

@ -28,7 +28,7 @@ train_dataloader = dict(
img_path='images/training', seg_map_path='annotations/training'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/chase_db1.py
View File

@ -32,7 +32,7 @@ train_dataloader = dict(
img_path='images/training', seg_map_path='annotations/training'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/cityscapes.py
View File

@ -28,7 +28,7 @@ train_dataloader = dict(
img_path='leftImg8bit/train', seg_map_path='gtFine/train'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/coco-stuff10k.py
View File

@ -29,7 +29,7 @@ train_dataloader = dict(
img_path='images/train2014', seg_map_path='annotations/train2014'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/coco-stuff164k.py
View File

@ -28,7 +28,7 @@ train_dataloader = dict(
img_path='images/train2017', seg_map_path='annotations/val2017'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/drive.py
View File

@ -31,7 +31,7 @@ train_dataloader = dict(
img_path='images/training', seg_map_path='annotations/training'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/hrf.py
View File

@ -31,7 +31,7 @@ train_dataloader = dict(
img_path='images/training', seg_map_path='annotations/training'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/isaid.py
View File

@ -35,7 +35,7 @@ train_dataloader = dict(
img_path='img_dir/train', seg_map_path='ann_dir/train'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/loveda.py
View File

@ -28,7 +28,7 @@ train_dataloader = dict(
img_path='img_dir/train', seg_map_path='ann_dir/train'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/pascal_context.py
View File

@ -32,7 +32,7 @@ train_dataloader = dict(
ann_file='ImageSets/SegmentationContext/train.txt',
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/pascal_context_59.py
View File

@ -32,7 +32,7 @@ train_dataloader = dict(
ann_file='ImageSets/SegmentationContext/train.txt',
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/pascal_voc12.py
View File

@ -29,7 +29,7 @@ train_dataloader = dict(
ann_file='ImageSets/Segmentation/train.txt',
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/potsdam.py
View File

@ -28,7 +28,7 @@ train_dataloader = dict(
img_path='img_dir/train', seg_map_path='ann_dir/train'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/stare.py
View File

@ -31,7 +31,7 @@ train_dataloader = dict(
img_path='images/training', seg_map_path='annotations/training'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/_base_/datasets/vaihingen.py
View File

@ -28,7 +28,7 @@ train_dataloader = dict(
img_path='img_dir/train', seg_map_path='ann_dir/train'),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=4,
batch_size=1,
num_workers=4,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=False),

2
configs/beit/upernet_beit-base_640x640_160k_ade20k_ms.py
View File

@ -9,5 +9,5 @@ test_pipeline = [
dict(type='Resize', scale=(2560, 640), keep_ratio=True),
dict(type='PackSegInputs'),
]
val_dataloader = dict(batch_size=2, dataset=dict(pipeline=test_pipeline))
val_dataloader = dict(batch_size=1, dataset=dict(pipeline=test_pipeline))
test_dataloader = val_dataloader

2
configs/beit/upernet_beit-base_8x2_640x640_160k_ade20k.py
View File

@ -37,5 +37,5 @@ param_scheduler = [
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader

2
configs/bisenetv1/bisenetv1_r101-d32_lr5e-3_4x4_512x512_160k_coco-stuff164k.py
View File

@ -31,5 +31,5 @@ param_scheduler = [
optimizer = dict(type='SGD', lr=0.005, momentum=0.9, weight_decay=0.0005)
optim_wrapper = dict(type='OptimWrapper', optimizer=optimizer)
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/bisenetv1/bisenetv1_r18-d32_4x4_1024x1024_160k_cityscapes.py
View File

@ -20,5 +20,5 @@ param_scheduler = [
optimizer = dict(type='SGD', lr=0.025, momentum=0.9, weight_decay=0.0005)
optim_wrapper = dict(type='OptimWrapper', optimizer=optimizer)
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/bisenetv1/bisenetv1_r18-d32_in1k-pre_4x4_1024x1024_160k_cityscapes.py
View File

@ -25,5 +25,5 @@ param_scheduler = [
optimizer = dict(type='SGD', lr=0.025, momentum=0.9, weight_decay=0.0005)
optim_wrapper = dict(type='OptimWrapper', optimizer=optimizer)
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/bisenetv1/bisenetv1_r18-d32_in1k-pre_4x8_1024x1024_160k_cityscapes.py
View File

@ -1,4 +1,4 @@
_base_ = './bisenetv1_r18-d32_in1k-pre_4x4_1024x1024_160k_cityscapes.py'
train_dataloader = dict(batch_size=8, num_workers=4)
val_dataloader = dict(batch_size=8, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/bisenetv1/bisenetv1_r18-d32_lr5e-3_4x4_512x512_160k_coco-stuff164k.py
View File

@ -26,5 +26,5 @@ param_scheduler = [
optimizer = dict(type='SGD', lr=0.005, momentum=0.9, weight_decay=0.0005)
optim_wrapper = dict(type='OptimWrapper', optimizer=optimizer)
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/bisenetv1/bisenetv1_r50-d32_4x4_1024x1024_160k_cityscapes.py
View File

@ -51,5 +51,5 @@ param_scheduler = [
optimizer = dict(type='SGD', lr=0.05, momentum=0.9, weight_decay=0.0005)
optim_wrapper = dict(type='OptimWrapper', optimizer=optimizer)
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/bisenetv1/bisenetv1_r50-d32_lr5e-3_4x4_512x512_160k_coco-stuff164k.py
View File

@ -31,5 +31,5 @@ param_scheduler = [
optimizer = dict(type='SGD', lr=0.005, momentum=0.9, weight_decay=0.0005)
optim_wrapper = dict(type='OptimWrapper', optimizer=optimizer)
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/bisenetv2/bisenetv2_fcn_4x4_1024x1024_160k_cityscapes.py
View File

@ -20,5 +20,5 @@ param_scheduler = [
optimizer = dict(type='SGD', lr=0.05, momentum=0.9, weight_decay=0.0005)
optim_wrapper = dict(type='OptimWrapper', optimizer=optimizer)
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/bisenetv2/bisenetv2_fcn_4x8_1024x1024_160k_cityscapes.py
View File

@ -20,5 +20,5 @@ param_scheduler = [
optimizer = dict(type='SGD', lr=0.05, momentum=0.9, weight_decay=0.0005)
optim_wrapper = dict(type='OptimWrapper', optimizer=optimizer)
train_dataloader = dict(batch_size=8, num_workers=4)
val_dataloader = dict(batch_size=8, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/bisenetv2/bisenetv2_fcn_ohem_4x4_1024x1024_160k_cityscapes.py
View File

@ -79,5 +79,5 @@ param_scheduler = [
optimizer = dict(type='SGD', lr=0.05, momentum=0.9, weight_decay=0.0005)
optim_wrapper = dict(type='OptimWrapper', optimizer=optimizer)
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/cgnet/cgnet_512x1024_60k_cityscapes.py
View File

@ -29,5 +29,5 @@ preprocess_cfg = dict(size=crop_size)
model = dict(preprocess_cfg=preprocess_cfg)
train_dataloader = dict(batch_size=8)
val_dataloader = dict(batch_size=8)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader

2
configs/cgnet/cgnet_680x680_60k_cityscapes.py
View File

@ -43,5 +43,5 @@ test_pipeline = [
train_dataloader = dict(
batch_size=8, num_workers=4, dataset=dict(pipeline=train_pipeline))
val_dataloader = dict(
batch_size=8, num_workers=4, dataset=dict(pipeline=test_pipeline))
batch_size=1, num_workers=4, dataset=dict(pipeline=test_pipeline))
test_dataloader = val_dataloader

2
configs/convnext/upernet_convnext_base_fp16_512x512_160k_ade20k.py
View File

@ -43,7 +43,7 @@ param_scheduler = [
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader
# fp16 settings
default_hooks = dict(

2
configs/convnext/upernet_convnext_base_fp16_640x640_160k_ade20k.py
View File

@ -58,7 +58,7 @@ param_scheduler = [
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader
# fp16 settings
default_hooks = dict(

2
configs/convnext/upernet_convnext_large_fp16_640x640_160k_ade20k.py
View File

@ -58,7 +58,7 @@ param_scheduler = [
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader
# fp16 settings
default_hooks = dict(

2
configs/convnext/upernet_convnext_small_fp16_512x512_160k_ade20k.py
View File

@ -57,7 +57,7 @@ param_scheduler = [
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader
# fp16 settings
default_hooks = dict(

2
configs/convnext/upernet_convnext_tiny_fp16_512x512_160k_ade20k.py
View File

@ -57,7 +57,7 @@ param_scheduler = [
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader
# fp16 settings
default_hooks = dict(

2
configs/convnext/upernet_convnext_xlarge_fp16_640x640_160k_ade20k.py
View File

@ -58,7 +58,7 @@ param_scheduler = [
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader
# fp16 settings
default_hooks = dict(

2
configs/dpt/dpt_vit-b16_512x512_160k_ade20k.py
View File

@ -39,5 +39,5 @@ param_scheduler = [
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/erfnet/erfnet_fcn_4x4_512x1024_160k_cityscapes.py
View File

@ -6,5 +6,5 @@ crop_size = (512, 1024)
preprocess_cfg = dict(size=crop_size)
model = dict(preprocess_cfg=preprocess_cfg)
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/fastfcn/fastfcn_r50-d32_jpu_aspp_4x4_512x1024_80k_cityscapes.py
View File

@ -1,5 +1,5 @@
# model settings
_base_ = './fastfcn_r50-d32_jpu_aspp_512x1024_80k_cityscapes.py'
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/fastfcn/fastfcn_r50-d32_jpu_enc_4x4_512x1024_80k_cityscapes.py
View File

@ -1,5 +1,5 @@
# model settings
_base_ = './fastfcn_r50-d32_jpu_enc_512x1024_80k_cityscapes.py'
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/fastfcn/fastfcn_r50-d32_jpu_psp_4x4_512x1024_80k_cityscapes.py
View File

@ -7,5 +7,5 @@ crop_size = (512, 1024)
preprocess_cfg = dict(size=crop_size)
model = dict(preprocess_cfg=preprocess_cfg)
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/fastscnn/fast_scnn_lr0.12_8x4_160k_cityscapes.py
View File

@ -7,7 +7,7 @@ preprocess_cfg = dict(size=crop_size)
model = dict(preprocess_cfg=preprocess_cfg)
# Re-config the data sampler.
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader
# Re-config the optimizer.

2
configs/knet/knet_s3_deeplabv3_r50-d8_8x2_512x512_adamw_80k_ade20k.py
View File

@ -102,5 +102,5 @@ param_scheduler = [
]
# In K-Net implementation we use batch size 2 per GPU as default
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/knet/knet_s3_fcn_r50-d8_8x2_512x512_adamw_80k_ade20k.py
View File

@ -102,5 +102,5 @@ param_scheduler = [
]
# In K-Net implementation we use batch size 2 per GPU as default
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/knet/knet_s3_pspnet_r50-d8_8x2_512x512_adamw_80k_ade20k.py
View File

@ -101,5 +101,5 @@ param_scheduler = [
]
# In K-Net implementation we use batch size 2 per GPU as default
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/knet/knet_s3_upernet_r50-d8_8x2_512x512_adamw_80k_ade20k.py
View File

@ -102,5 +102,5 @@ param_scheduler = [
]
# In K-Net implementation we use batch size 2 per GPU as default
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/knet/knet_s3_upernet_swin-l_8x2_512x512_adamw_80k_ade20k.py
View File

@ -17,5 +17,5 @@ model = dict(
auxiliary_head=dict(in_channels=768))
# In K-Net implementation we use batch size 2 per GPU as default
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/knet/knet_s3_upernet_swin-l_8x2_640x640_adamw_80k_ade20k.py
View File

@ -39,5 +39,5 @@ val_dataloader = dict(dataset=dict(pipeline=test_pipeline))
test_dataloader = val_dataloader
# In K-Net implementation we use batch size 2 per GPU as default
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/knet/knet_s3_upernet_swin-t_8x2_512x512_adamw_80k_ade20k.py
View File

@ -68,5 +68,5 @@ param_scheduler = [
]
# In K-Net implementation we use batch size 2 per GPU as default
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/mae/upernet_mae-base_fp16_512x512_160k_ade20k_ms.py
View File

@ -9,5 +9,5 @@ test_pipeline = [
dict(type='Resize', scale=(2048, 512), keep_ratio=True),
dict(type='PackSegInputs')
]
val_dataloader = dict(batch_size=2, dataset=dict(pipeline=test_pipeline))
val_dataloader = dict(batch_size=1, dataset=dict(pipeline=test_pipeline))
test_dataloader = val_dataloader

2
configs/mae/upernet_mae-base_fp16_8x2_512x512_160k_ade20k.py
View File

@ -55,5 +55,5 @@ fp16 = dict(loss_scale='dynamic')
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader

2
configs/mobilenet_v3/lraspp_m-v3-d8_512x1024_320k_cityscapes.py
View File

@ -10,7 +10,7 @@ model = dict(
# Re-config the data sampler.
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader
runner = dict(type='IterBasedRunner', max_iters=320000)

2
configs/mobilenet_v3/lraspp_m-v3-d8_scratch_512x1024_320k_cityscapes.py
View File

@ -7,7 +7,7 @@ preprocess_cfg = dict(size=crop_size)
# Re-config the data sampler.
model = dict(preprocess_cfg=preprocess_cfg)
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader
runner = dict(type='IterBasedRunner', max_iters=320000)

2
configs/segformer/segformer_mit-b0_512x512_160k_ade20k.py
View File

@ -40,5 +40,5 @@ param_scheduler = [
)
]
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

4
configs/segformer/segformer_mit-b0_8x1_1024x1024_160k_cityscapes.py
View File

@ -42,6 +42,6 @@ param_scheduler = [
)
]
train_dataloader = dict(batch_size=1, num_workers=1)
val_dataloader = dict(batch_size=1, num_workers=1)
train_dataloader = dict(batch_size=1, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/setr/setr_mla_512x512_160k_b16_ade20k.py
View File

@ -2,5 +2,5 @@ _base_ = ['./setr_mla_512x512_160k_b8_ade20k.py']
# num_gpus: 8 -> batch_size: 16
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader

2
configs/setr/setr_naive_512x512_160k_b16_ade20k.py
View File

@ -68,5 +68,5 @@ optimizer = dict(
# num_gpus: 8 -> batch_size: 16
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader

2
configs/setr/setr_pup_512x512_160k_b16_ade20k.py
View File

@ -68,5 +68,5 @@ optimizer = dict(
# num_gpus: 8 -> batch_size: 16
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader

2
configs/stdc/stdc1_512x1024_80k_cityscapes.py
View File

@ -17,5 +17,5 @@ param_scheduler = [
)
]
train_dataloader = dict(batch_size=12, num_workers=4)
val_dataloader = dict(batch_size=12, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/swin/upernet_swin_tiny_patch4_window7_512x512_160k_ade20k_pretrain_224x224_1K.py
View File

@ -53,5 +53,5 @@ param_scheduler = [
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader

2
configs/twins/twins_pcpvt-b_uperhead_8x2_512x512_160k_ade20k.py
View File

@ -9,5 +9,5 @@ model = dict(
drop_path_rate=0.3))
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/twins/twins_pcpvt-l_uperhead_8x2_512x512_160k_ade20k.py
View File

@ -9,5 +9,5 @@ model = dict(
drop_path_rate=0.3))
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/twins/twins_svt-s_uperhead_8x2_512x512_160k_ade20k.py
View File

@ -50,5 +50,5 @@ param_scheduler = [
]
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/unet/fcn_unet_s5-d16_4x4_512x1024_160k_cityscapes.py
View File

@ -12,5 +12,5 @@ model = dict(
train_cfg=dict(),
test_cfg=dict(mode='whole'))
train_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=4, num_workers=4)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

2
configs/vit/upernet_vit-b16_ln_mln_512x512_160k_ade20k.py
View File

@ -46,5 +46,5 @@ param_scheduler = [
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader

2
configs/vit/upernet_vit-b16_mln_512x512_160k_ade20k.py
View File

@ -45,5 +45,5 @@ param_scheduler = [
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader

2
configs/vit/upernet_vit-b16_mln_512x512_80k_ade20k.py
View File

@ -45,5 +45,5 @@ param_scheduler = [
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=2)
val_dataloader = dict(batch_size=1)
test_dataloader = val_dataloader