diff --git a/configs/fastscnn/fast_scnn_4x3_1000e_cityscapes.py b/configs/fastscnn/fast_scnn_4x3_1000e_cityscapes.py
deleted file mode 100644
index 7583cb6f2..000000000
--- a/configs/fastscnn/fast_scnn_4x3_1000e_cityscapes.py
+++ /dev/null
@@ -1,62 +0,0 @@
-_base_ = [
-    '../_base_/models/fast_scnn.py', '../_base_/datasets/cityscapes.py',
-    '../_base_/default_runtime.py'
-]
-crop_size = (512, 1024)
-cudnn_benchmark = True
-# model training and testing settings
-train_cfg = dict()
-test_cfg = dict(mode='whole')
-
-# Here: What is parameter 'with_seg'?
-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', to_float32=True),
-    dict(type='LoadAnnotations'),   # with_seg=True
-    dict(type='Resize', img_scale=(2048, 1024), ratio_range=(0.5, 2.0)),
-    dict(type='RandomFlip', flip_ratio=0.5),
-    dict(type='PhotoMetricDistortion'),
-    dict(type='Normalize', **img_norm_cfg),
-    dict(type='RandomCrop', crop_size=crop_size),
-    dict(type='DefaultFormatBundle'),
-    dict(type='Collect', keys=['img', 'gt_semantic_seg']),
-]
-test_pipeline = [
-    dict(type='LoadImageFromFile'),
-    dict(
-        type='MultiScaleFlipAug',
-        img_scale=(2048, 1024),
-        # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75],
-        flip=False,
-        transforms=[
-            dict(type='Resize', keep_ratio=True),
-            dict(type='RandomFlip'),
-            dict(type='Normalize', **img_norm_cfg),
-            dict(type='ImageToTensor', keys=['img']),
-            dict(type='Collect', keys=['img']),
-        ])
-]
-data = dict(
-    samples_per_gpu=3,
-    workers_per_gpu=3,
-    train=dict(pipeline=train_pipeline),
-    val=dict(pipeline=test_pipeline),
-    test=dict(pipeline=test_pipeline))
-
-# optimizer
-optimizer = dict(type='SGD', lr=0.045, momentum=0.9, weight_decay=4e-5)
-optimizer_config = dict()
-# learning policy
-lr_config = dict(
-    policy='poly',
-    power=0.9,
-    by_epoch=False,
-)
-# runtime settings
-total_epochs = 247000
-evaluation = dict(interval=1000, metric='mIoU')
-checkpoint_config = dict(interval=1000)
-
-# log config: log by iter.
-log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook', by_epoch=False)])
diff --git a/configs/fastscnn/fast_scnn_4x3_450k_lr0.08_cityscapes.py b/configs/fastscnn/fast_scnn_4x3_450k_lr0.08_cityscapes.py
deleted file mode 100644
index f562f0d76..000000000
--- a/configs/fastscnn/fast_scnn_4x3_450k_lr0.08_cityscapes.py
+++ /dev/null
@@ -1,64 +0,0 @@
-_base_ = [
-    '../_base_/models/fast_scnn.py', '../_base_/datasets/cityscapes.py',
-    '../_base_/default_runtime.py'
-]
-crop_size = (512, 1024)
-cudnn_benchmark = True
-# model training and testing settings
-train_cfg = dict()
-test_cfg = dict(mode='whole')
-
-# Here: What is parameter 'with_seg'?
-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', to_float32=True),
-    dict(type='LoadAnnotations'),   # with_seg=True
-    dict(type='Resize', img_scale=(2048, 1024), ratio_range=(0.5, 2.0)),
-    dict(type='RandomFlip', flip_ratio=0.5),
-    dict(type='PhotoMetricDistortion'),
-    dict(type='Normalize', **img_norm_cfg),
-    dict(type='RandomCrop', crop_size=crop_size),
-    dict(type='DefaultFormatBundle'),
-    dict(type='Collect', keys=['img', 'gt_semantic_seg']),
-]
-test_pipeline = [
-    dict(type='LoadImageFromFile'),
-    dict(
-        type='MultiScaleFlipAug',
-        img_scale=(2048, 1024),
-        # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75],
-        flip=False,
-        transforms=[
-            dict(type='Resize', keep_ratio=True),
-            dict(type='RandomFlip'),
-            dict(type='Normalize', **img_norm_cfg),
-            dict(type='ImageToTensor', keys=['img']),
-            dict(type='Collect', keys=['img']),
-        ])
-]
-data = dict(
-    samples_per_gpu=3,
-    workers_per_gpu=3,
-    train=dict(pipeline=train_pipeline),
-    val=dict(pipeline=test_pipeline),
-    test=dict(pipeline=test_pipeline))
-
-# optimizer
-optimizer = dict(type='SGD', lr=0.08, momentum=0.9, weight_decay=4e-5)
-optimizer_config = dict()
-# learning policy
-lr_config = dict(
-    policy='poly',
-    power=0.9,
-    by_epoch=False,
-)
-# runtime settings
-# total_epochs = 1000
-total_iters = 450000
-evaluation = dict(interval=2000, metric='mIoU')
-checkpoint_config = dict(interval=2000)
-
-# log config: log by iter.
-log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook', by_epoch=False)])
-
diff --git a/configs/fastscnn/fast_scnn_4x8_100k_lr0.045_power1.2_cityscapes.py b/configs/fastscnn/fast_scnn_4x8_100k_lr0.045_power1.2_cityscapes.py
deleted file mode 100644
index 078fa94c5..000000000
--- a/configs/fastscnn/fast_scnn_4x8_100k_lr0.045_power1.2_cityscapes.py
+++ /dev/null
@@ -1,64 +0,0 @@
-_base_ = [
-    '../_base_/models/fast_scnn.py', '../_base_/datasets/cityscapes.py',
-    '../_base_/default_runtime.py'
-]
-crop_size = (512, 1024)
-cudnn_benchmark = True
-# model training and testing settings
-train_cfg = dict()
-test_cfg = dict(mode='whole')
-
-# Here: What is parameter 'with_seg'?
-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', to_float32=True),
-    dict(type='LoadAnnotations'),   # with_seg=True
-    dict(type='Resize', img_scale=(2048, 1024), ratio_range=(0.5, 2.0)),
-    dict(type='RandomFlip', flip_ratio=0.5),
-    dict(type='PhotoMetricDistortion'),
-    dict(type='Normalize', **img_norm_cfg),
-    dict(type='RandomCrop', crop_size=crop_size),
-    dict(type='DefaultFormatBundle'),
-    dict(type='Collect', keys=['img', 'gt_semantic_seg']),
-]
-test_pipeline = [
-    dict(type='LoadImageFromFile'),
-    dict(
-        type='MultiScaleFlipAug',
-        img_scale=(2048, 1024),
-        # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75],
-        flip=False,
-        transforms=[
-            dict(type='Resize', keep_ratio=True),
-            dict(type='RandomFlip'),
-            dict(type='Normalize', **img_norm_cfg),
-            dict(type='ImageToTensor', keys=['img']),
-            dict(type='Collect', keys=['img']),
-        ])
-]
-data = dict(
-    samples_per_gpu=8,
-    workers_per_gpu=4,
-    train=dict(pipeline=train_pipeline),
-    val=dict(pipeline=test_pipeline),
-    test=dict(pipeline=test_pipeline))
-
-# optimizer
-optimizer = dict(type='SGD', lr=0.045, momentum=0.9, weight_decay=4e-5)
-optimizer_config = dict()
-# learning policy
-lr_config = dict(
-    policy='poly',
-    power=1.2,
-    by_epoch=False,
-)
-# runtime settings
-# total_epochs = 1000
-total_iters = 100000
-evaluation = dict(interval=2000, metric='mIoU')
-checkpoint_config = dict(interval=2000)
-
-# log config: log by iter.
-log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook', by_epoch=False)])
-
diff --git a/configs/fastscnn/fast_scnn_4x8_100k_lr0.08_cityscapes.py b/configs/fastscnn/fast_scnn_4x8_100k_lr0.08_cityscapes.py
deleted file mode 100644
index 10f56e093..000000000
--- a/configs/fastscnn/fast_scnn_4x8_100k_lr0.08_cityscapes.py
+++ /dev/null
@@ -1,64 +0,0 @@
-_base_ = [
-    '../_base_/models/fast_scnn.py', '../_base_/datasets/cityscapes.py',
-    '../_base_/default_runtime.py'
-]
-crop_size = (512, 1024)
-cudnn_benchmark = True
-# model training and testing settings
-train_cfg = dict()
-test_cfg = dict(mode='whole')
-
-# Here: What is parameter 'with_seg'?
-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', to_float32=True),
-    dict(type='LoadAnnotations'),   # with_seg=True
-    dict(type='Resize', img_scale=(2048, 1024), ratio_range=(0.5, 2.0)),
-    dict(type='RandomFlip', flip_ratio=0.5),
-    dict(type='PhotoMetricDistortion'),
-    dict(type='Normalize', **img_norm_cfg),
-    dict(type='RandomCrop', crop_size=crop_size),
-    dict(type='DefaultFormatBundle'),
-    dict(type='Collect', keys=['img', 'gt_semantic_seg']),
-]
-test_pipeline = [
-    dict(type='LoadImageFromFile'),
-    dict(
-        type='MultiScaleFlipAug',
-        img_scale=(2048, 1024),
-        # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75],
-        flip=False,
-        transforms=[
-            dict(type='Resize', keep_ratio=True),
-            dict(type='RandomFlip'),
-            dict(type='Normalize', **img_norm_cfg),
-            dict(type='ImageToTensor', keys=['img']),
-            dict(type='Collect', keys=['img']),
-        ])
-]
-data = dict(
-    samples_per_gpu=8,
-    workers_per_gpu=4,
-    train=dict(pipeline=train_pipeline),
-    val=dict(pipeline=test_pipeline),
-    test=dict(pipeline=test_pipeline))
-
-# optimizer
-optimizer = dict(type='SGD', lr=0.08, momentum=0.9, weight_decay=4e-5)
-optimizer_config = dict()
-# learning policy
-lr_config = dict(
-    policy='poly',
-    power=0.9,
-    by_epoch=False,
-)
-# runtime settings
-# total_epochs = 1000
-total_iters = 100000
-evaluation = dict(interval=2000, metric='mIoU')
-checkpoint_config = dict(interval=2000)
-
-# log config: log by iter.
-log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook', by_epoch=False)])
-
diff --git a/configs/fastscnn/fast_scnn_4x8_100k_lr0.12_cityscapes.py b/configs/fastscnn/fast_scnn_4x8_100k_lr0.12_cityscapes.py
deleted file mode 100644
index 9ff49a9db..000000000
--- a/configs/fastscnn/fast_scnn_4x8_100k_lr0.12_cityscapes.py
+++ /dev/null
@@ -1,64 +0,0 @@
-_base_ = [
-    '../_base_/models/fast_scnn.py', '../_base_/datasets/cityscapes.py',
-    '../_base_/default_runtime.py'
-]
-crop_size = (512, 1024)
-cudnn_benchmark = True
-# model training and testing settings
-train_cfg = dict()
-test_cfg = dict(mode='whole')
-
-# Here: What is parameter 'with_seg'?
-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', to_float32=True),
-    dict(type='LoadAnnotations'),   # with_seg=True
-    dict(type='Resize', img_scale=(2048, 1024), ratio_range=(0.5, 2.0)),
-    dict(type='RandomFlip', flip_ratio=0.5),
-    dict(type='PhotoMetricDistortion'),
-    dict(type='Normalize', **img_norm_cfg),
-    dict(type='RandomCrop', crop_size=crop_size),
-    dict(type='DefaultFormatBundle'),
-    dict(type='Collect', keys=['img', 'gt_semantic_seg']),
-]
-test_pipeline = [
-    dict(type='LoadImageFromFile'),
-    dict(
-        type='MultiScaleFlipAug',
-        img_scale=(2048, 1024),
-        # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75],
-        flip=False,
-        transforms=[
-            dict(type='Resize', keep_ratio=True),
-            dict(type='RandomFlip'),
-            dict(type='Normalize', **img_norm_cfg),
-            dict(type='ImageToTensor', keys=['img']),
-            dict(type='Collect', keys=['img']),
-        ])
-]
-data = dict(
-    samples_per_gpu=8,
-    workers_per_gpu=4,
-    train=dict(pipeline=train_pipeline),
-    val=dict(pipeline=test_pipeline),
-    test=dict(pipeline=test_pipeline))
-
-# optimizer
-optimizer = dict(type='SGD', lr=0.12, momentum=0.9, weight_decay=4e-5)
-optimizer_config = dict()
-# learning policy
-lr_config = dict(
-    policy='poly',
-    power=0.9,
-    by_epoch=False,
-)
-# runtime settings
-# total_epochs = 1000
-total_iters = 100000
-evaluation = dict(interval=2000, metric='mIoU')
-checkpoint_config = dict(interval=2000)
-
-# log config: log by iter.
-log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook', by_epoch=False)])
-
diff --git a/configs/fastscnn/fast_scnn_4x8_80k_cityscapes.py b/configs/fastscnn/fast_scnn_4x8_80k_cityscapes.py
deleted file mode 100644
index dd6ac7f15..000000000
--- a/configs/fastscnn/fast_scnn_4x8_80k_cityscapes.py
+++ /dev/null
@@ -1,61 +0,0 @@
-_base_ = [
-    '../_base_/models/fast_scnn.py', '../_base_/datasets/cityscapes.py',
-    '../_base_/default_runtime.py'
-]
-crop_size = (512, 1024)
-cudnn_benchmark = True
-# model training and testing settings
-train_cfg = dict()
-test_cfg = dict(mode='whole')
-
-# Here: What is parameter 'with_seg'?
-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', to_float32=True),
-    dict(type='LoadAnnotations'),   # with_seg=True
-    dict(type='Resize', img_scale=(2048, 1024), ratio_range=(0.5, 2.0)),
-    dict(type='RandomFlip', flip_ratio=0.5),
-    dict(type='PhotoMetricDistortion'),
-    dict(type='Normalize', **img_norm_cfg),
-    dict(type='RandomCrop', crop_size=crop_size),
-    dict(type='DefaultFormatBundle'),
-    dict(type='Collect', keys=['img', 'gt_semantic_seg']),
-]
-test_pipeline = [
-    dict(type='LoadImageFromFile'),
-    dict(
-        type='MultiScaleFlipAug',
-        img_scale=(2048, 1024),
-        # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75],
-        flip=False,
-        transforms=[
-            dict(type='Resize', keep_ratio=True),
-            dict(type='RandomFlip'),
-            dict(type='Normalize', **img_norm_cfg),
-            dict(type='ImageToTensor', keys=['img']),
-            dict(type='Collect', keys=['img']),
-        ])
-]
-data = dict(
-    samples_per_gpu=8,
-    workers_per_gpu=4,
-    train=dict(pipeline=train_pipeline),
-    val=dict(pipeline=test_pipeline),
-    test=dict(pipeline=test_pipeline))
-
-# optimizer
-optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=4e-5)
-optimizer_config = dict()
-# learning policy
-lr_config = dict(
-    policy='poly',
-    power=0.9,
-    by_epoch=False,
-)
-# runtime settings
-# total_epochs = 1000
-total_iters = 100000
-evaluation = dict(interval=2000, metric='mIoU')
-checkpoint_config = dict(interval=2000)
-
diff --git a/configs/fastscnn/fast_scnn_4x8_80k_lr0.045_cityscapes.py b/configs/fastscnn/fast_scnn_4x8_80k_lr0.045_cityscapes.py
deleted file mode 100644
index f6fd1729e..000000000
--- a/configs/fastscnn/fast_scnn_4x8_80k_lr0.045_cityscapes.py
+++ /dev/null
@@ -1,64 +0,0 @@
-_base_ = [
-    '../_base_/models/fast_scnn.py', '../_base_/datasets/cityscapes.py',
-    '../_base_/default_runtime.py'
-]
-crop_size = (512, 1024)
-cudnn_benchmark = True
-# model training and testing settings
-train_cfg = dict()
-test_cfg = dict(mode='whole')
-
-# Here: What is parameter 'with_seg'?
-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', to_float32=True),
-    dict(type='LoadAnnotations'),   # with_seg=True
-    dict(type='Resize', img_scale=(2048, 1024), ratio_range=(0.5, 2.0)),
-    dict(type='RandomFlip', flip_ratio=0.5),
-    dict(type='PhotoMetricDistortion'),
-    dict(type='Normalize', **img_norm_cfg),
-    dict(type='RandomCrop', crop_size=crop_size),
-    dict(type='DefaultFormatBundle'),
-    dict(type='Collect', keys=['img', 'gt_semantic_seg']),
-]
-test_pipeline = [
-    dict(type='LoadImageFromFile'),
-    dict(
-        type='MultiScaleFlipAug',
-        img_scale=(2048, 1024),
-        # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75],
-        flip=False,
-        transforms=[
-            dict(type='Resize', keep_ratio=True),
-            dict(type='RandomFlip'),
-            dict(type='Normalize', **img_norm_cfg),
-            dict(type='ImageToTensor', keys=['img']),
-            dict(type='Collect', keys=['img']),
-        ])
-]
-data = dict(
-    samples_per_gpu=8,
-    workers_per_gpu=4,
-    train=dict(pipeline=train_pipeline),
-    val=dict(pipeline=test_pipeline),
-    test=dict(pipeline=test_pipeline))
-
-# optimizer
-optimizer = dict(type='SGD', lr=0.045, momentum=0.9, weight_decay=4e-5)
-optimizer_config = dict()
-# learning policy
-lr_config = dict(
-    policy='poly',
-    power=0.9,
-    by_epoch=False,
-)
-# runtime settings
-# total_epochs = 1000
-total_iters = 100000
-evaluation = dict(interval=2000, metric='mIoU')
-checkpoint_config = dict(interval=2000)
-
-# log config: log by iter.
-log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook', by_epoch=False)])
-
diff --git a/configs/fastscnn/fast_scnn_4x8_80k_lr0.08_cityscapes.py b/configs/fastscnn/fast_scnn_4x8_80k_lr0.08_cityscapes.py
deleted file mode 100644
index aec2ea0a2..000000000
--- a/configs/fastscnn/fast_scnn_4x8_80k_lr0.08_cityscapes.py
+++ /dev/null
@@ -1,64 +0,0 @@
-_base_ = [
-    '../_base_/models/fast_scnn.py', '../_base_/datasets/cityscapes.py',
-    '../_base_/default_runtime.py'
-]
-crop_size = (512, 1024)
-cudnn_benchmark = True
-# model training and testing settings
-train_cfg = dict()
-test_cfg = dict(mode='whole')
-
-# Here: What is parameter 'with_seg'?
-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', to_float32=True),
-    dict(type='LoadAnnotations'),   # with_seg=True
-    dict(type='Resize', img_scale=(2048, 1024), ratio_range=(0.5, 2.0)),
-    dict(type='RandomFlip', flip_ratio=0.5),
-    dict(type='PhotoMetricDistortion'),
-    dict(type='Normalize', **img_norm_cfg),
-    dict(type='RandomCrop', crop_size=crop_size),
-    dict(type='DefaultFormatBundle'),
-    dict(type='Collect', keys=['img', 'gt_semantic_seg']),
-]
-test_pipeline = [
-    dict(type='LoadImageFromFile'),
-    dict(
-        type='MultiScaleFlipAug',
-        img_scale=(2048, 1024),
-        # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75],
-        flip=False,
-        transforms=[
-            dict(type='Resize', keep_ratio=True),
-            dict(type='RandomFlip'),
-            dict(type='Normalize', **img_norm_cfg),
-            dict(type='ImageToTensor', keys=['img']),
-            dict(type='Collect', keys=['img']),
-        ])
-]
-data = dict(
-    samples_per_gpu=8,
-    workers_per_gpu=4,
-    train=dict(pipeline=train_pipeline),
-    val=dict(pipeline=test_pipeline),
-    test=dict(pipeline=test_pipeline))
-
-# optimizer
-optimizer = dict(type='SGD', lr=0.08, momentum=0.9, weight_decay=4e-5)
-optimizer_config = dict()
-# learning policy
-lr_config = dict(
-    policy='poly',
-    power=0.9,
-    by_epoch=False,
-)
-# runtime settings
-# total_epochs = 1000
-total_iters = 80000
-evaluation = dict(interval=2000, metric='mIoU')
-checkpoint_config = dict(interval=2000)
-
-# log config: log by iter.
-log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook', by_epoch=False)])
-