| | --- |
| | tags: |
| | - image-classification |
| | - timm |
| | - transformers |
| | library_name: timm |
| | license: apache-2.0 |
| | --- |
| | # Model card for resnext101d_32x4d |
| | |
| | This repo is provided with ResNeXt101d_32x4d trained by timm. |
| |
|
| | ## Usage |
| |
|
| | ```python |
| | model = timm.create_model("hf_hub:mjun0812/resnext101d_32x4d", pretrained=True) |
| | ``` |
| |
|
| | ## Detail |
| |
|
| | ImageNet-1K results |
| |
|
| | | top1 | top5 | |
| | | -- | -- | |
| | | 80.9695 | 95.1044 | |
| |
|
| | config |
| |
|
| | ```python |
| | @register_model |
| | def resnext101d_32x4d(pretrained: bool = False, **kwargs) -> ResNet: |
| | """Constructs a ResNeXt101d 32x4d model.""" |
| | model_args = dict( |
| | block=Bottleneck, |
| | layers=(3, 4, 23, 3), |
| | cardinality=32, |
| | base_width=4, |
| | stem_width=32, |
| | stem_type="deep", |
| | avg_down=True, |
| | ) |
| | return _create_resnet("resnext101d_32x4d", pretrained, **dict(model_args, **kwargs)) |
| | ``` |
| |
|
| | training command |
| |
|
| | ```bash |
| | torchrun train.py \ |
| | --data-dir ~/workspace/dataset/ImageNet --model resnext101d_32x4d --lr 0.6 --warmup-epochs 5 --epochs 240 \ |
| | --weight-decay 1e-4 --sched cosine --reprob 0.4 --recount 3 --remode pixel --aa rand-m7-mstd0.5-inc1 -b 256 -j 6 --amp --dist-bn reduce |
| | ``` |
| |
|
