mmpretrain/configs/milan/metafile.yml

60 lines
2.3 KiB
YAML

Collections:
- Name: MILAN
Metadata:
Training Data: ImageNet-1k
Training Techniques:
- AdamW
Training Resources: 16x A100-80G GPUs
Architecture:
- ViT
Paper:
Title: 'MILAN: Masked Image Pretraining on Language Assisted Representation'
URL: https://arxiv.org/pdf/2208.06049
README: configs/milan/README.md
Models:
- Name: milan_vit-base-p16_16xb256-amp-coslr-400e_in1k
Metadata:
Epochs: 400
Batch Size: 4096
FLOPs: 17581972224
Parameters: 111907584
Training Data: ImageNet-1k
In Collection: MILAN
Results: null
Weights: https://download.openmmlab.com/mmselfsup/1.x/milan/milan_vit-base-p16_16xb256-amp-coslr-400e_in1k/milan_vit-base-p16_16xb256-amp-coslr-400e_in1k_20221129-180922e8.pth
Config: configs/milan/milan_vit-base-p16_16xb256-amp-coslr-400e_in1k.py
Downstream:
- vit-base-p16_milan-pre_8xb128-coslr-100e_in1k
- vit-base-p16_milan-pre_8xb2048-linear-coslr-100e_in1k
- Name: vit-base-p16_milan-pre_8xb128-coslr-100e_in1k
Metadata:
Epochs: 100
Batch Size: 1024
FLOPs: 17581215744
Parameters: 86566120
Training Data: ImageNet-1k
In Collection: MILAN
Results:
- Task: Image Classification
Dataset: ImageNet-1k
Metrics:
Top 1 Accuracy: 85.3
Weights: https://download.openmmlab.com/mmselfsup/1.x/milan/milan_vit-base-p16_16xb256-amp-coslr-400e_in1k/vit-base-p16_ft-8xb128-coslr-100e_in1k/vit-base-p16_ft-8xb128-coslr-100e_in1k-milan_20221129-74ac94fa.pth
Config: configs/milan/benchmarks/vit-base-p16_8xb128-coslr-100e_in1k.py
- Name: vit-base-p16_milan-pre_8xb2048-linear-coslr-100e_in1k
Metadata:
Epochs: 100
Batch Size: 16384
FLOPs: 17581972992
Parameters: 86567656
Training Data: ImageNet-1k
In Collection: MILAN
Results:
- Task: Image Classification
Dataset: ImageNet-1k
Metrics:
Top 1 Accuracy: 78.9
Weights: https://download.openmmlab.com/mmselfsup/1.x/milan/milan_vit-base-p16_16xb256-amp-coslr-400e_in1k/vit-base-p16_linear-8xb2048-coslr-100e_in1k/vit-base-p16_linear-8xb2048-coslr-100e_in1k_20221129-03f26f85.pth
Config: configs/milan/benchmarks/vit-base-p16_8xb2048-linear-coslr-100e_in1k.py