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
mmsegmentation/configs/convnext/convnext.yml

134 lines
4.4 KiB
YAML
Raw Normal View History

[Feature] Support ConvNext (#1216) * upload original backbone and configs * ConvNext Refactor * ConvNext Refactor * convnext customization refactor with mmseg style * convnext customization refactor with mmseg style * add ade20k_640x640.py * upload files for training * delete dist_optimizer_hook and remove layer_decay_optimizer_constructor * check max(out_indices) < num_stages * add unittest * fix lint error * use MMClassification backbone * fix bugs in base_1k * add mmcls in requirements/mminstall.txt * add mmcls in requirements/mminstall.txt * fix drop_path_rate and layer_scale_init_value * use logger.info instead of print * add mmcls in runtime.txt * fix f string && delete * add doctring in LearningRateDecayOptimizerConstructor and fix mmcls version in requirements * fix typo in LearningRateDecayOptimizerConstructor * use ConvNext models in unit test for LearningRateDecayOptimizerConstructor * add unit test * fix typo * fix typo * add layer_wise and fix redundant backbone.downsample_norm in it * fix unit test * give a ground truth lr_scale and weight_decay * upload models and readme * delete 'backbone.stem_norm' and 'backbone.downsample_norm' in get_num_layer() * fix unit test and use mmcls url * update md2yml.py and metafile * fix typo
2022-03-04 15:52:01 +08:00
Models:
- Name: upernet_convnext_tiny_fp16_512x512_160k_ade20k
In Collection: UperNet
Metadata:
backbone: ConvNeXt-T
crop size: (512,512)
lr schd: 160000
inference time (ms/im):
- value: 50.25
hardware: V100
backend: PyTorch
batch size: 1
mode: FP16
resolution: (512,512)
Training Memory (GB): 4.23
Results:
- Task: Semantic Segmentation
Dataset: ADE20K
Metrics:
mIoU: 46.11
mIoU(ms+flip): 46.62
Config: configs/convnext/upernet_convnext_tiny_fp16_512x512_160k_ade20k.py
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/convnext/upernet_convnext_tiny_fp16_512x512_160k_ade20k/upernet_convnext_tiny_fp16_512x512_160k_ade20k_20220227_124553-cad485de.pth
- Name: upernet_convnext_small_fp16_512x512_160k_ade20k
In Collection: UperNet
Metadata:
backbone: ConvNeXt-S
crop size: (512,512)
lr schd: 160000
inference time (ms/im):
- value: 65.88
hardware: V100
backend: PyTorch
batch size: 1
mode: FP16
resolution: (512,512)
Training Memory (GB): 5.16
Results:
- Task: Semantic Segmentation
Dataset: ADE20K
Metrics:
mIoU: 48.56
mIoU(ms+flip): 49.02
Config: configs/convnext/upernet_convnext_small_fp16_512x512_160k_ade20k.py
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/convnext/upernet_convnext_small_fp16_512x512_160k_ade20k/upernet_convnext_small_fp16_512x512_160k_ade20k_20220227_131208-1b1e394f.pth
- Name: upernet_convnext_base_fp16_512x512_160k_ade20k
In Collection: UperNet
Metadata:
backbone: ConvNeXt-B
crop size: (512,512)
lr schd: 160000
inference time (ms/im):
- value: 69.4
hardware: V100
backend: PyTorch
batch size: 1
mode: FP16
resolution: (512,512)
Training Memory (GB): 6.33
Results:
- Task: Semantic Segmentation
Dataset: ADE20K
Metrics:
mIoU: 48.71
mIoU(ms+flip): 49.54
Config: configs/convnext/upernet_convnext_base_fp16_512x512_160k_ade20k.py
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/convnext/upernet_convnext_base_fp16_512x512_160k_ade20k/upernet_convnext_base_fp16_512x512_160k_ade20k_20220227_181227-02a24fc6.pth
- Name: upernet_convnext_base_fp16_640x640_160k_ade20k
In Collection: UperNet
Metadata:
backbone: ConvNeXt-B
crop size: (640,640)
lr schd: 160000
inference time (ms/im):
- value: 91.91
hardware: V100
backend: PyTorch
batch size: 1
mode: FP16
resolution: (640,640)
Training Memory (GB): 8.53
Results:
- Task: Semantic Segmentation
Dataset: ADE20K
Metrics:
mIoU: 52.13
mIoU(ms+flip): 52.66
Config: configs/convnext/upernet_convnext_base_fp16_640x640_160k_ade20k.py
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/convnext/upernet_convnext_base_fp16_640x640_160k_ade20k/upernet_convnext_base_fp16_640x640_160k_ade20k_20220227_182859-9280e39b.pth
- Name: upernet_convnext_large_fp16_640x640_160k_ade20k
In Collection: UperNet
Metadata:
backbone: ConvNeXt-L
crop size: (640,640)
lr schd: 160000
inference time (ms/im):
- value: 130.04
hardware: V100
backend: PyTorch
batch size: 1
mode: FP16
resolution: (640,640)
Training Memory (GB): 12.08
Results:
- Task: Semantic Segmentation
Dataset: ADE20K
Metrics:
mIoU: 53.16
mIoU(ms+flip): 53.38
Config: configs/convnext/upernet_convnext_large_fp16_640x640_160k_ade20k.py
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/convnext/upernet_convnext_large_fp16_640x640_160k_ade20k/upernet_convnext_large_fp16_640x640_160k_ade20k_20220226_040532-e57aa54d.pth
- Name: upernet_convnext_xlarge_fp16_640x640_160k_ade20k
In Collection: UperNet
Metadata:
backbone: ConvNeXt-XL
crop size: (640,640)
lr schd: 160000
inference time (ms/im):
- value: 157.98
hardware: V100
backend: PyTorch
batch size: 1
mode: FP16
resolution: (640,640)
Training Memory (GB): 26.16
Results:
- Task: Semantic Segmentation
Dataset: ADE20K
Metrics:
mIoU: 53.58
mIoU(ms+flip): 54.11
Config: configs/convnext/upernet_convnext_xlarge_fp16_640x640_160k_ade20k.py
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/convnext/upernet_convnext_xlarge_fp16_640x640_160k_ade20k/upernet_convnext_xlarge_fp16_640x640_160k_ade20k_20220226_080344-95fc38c2.pth
Reference in New Issue Copy Permalink