36 lines
1.1 KiB
YAML
36 lines
1.1 KiB
YAML
Collections:
|
|
- Name: FP16
|
|
Metadata:
|
|
Training Data: ImageNet-1k
|
|
Training Techniques:
|
|
- SGD with Momentum
|
|
- Weight Decay
|
|
- Mixed Precision Training
|
|
Training Resources: 8x V100 GPUs
|
|
Paper:
|
|
URL: https://arxiv.org/abs/1710.03740
|
|
Title: Mixed Precision Training
|
|
README: configs/fp16/README.md
|
|
Code:
|
|
URL: https://github.com/open-mmlab/mmclassification/blob/a41cb2fa938d957101cc446e271486206188bf5b/mmcls/core/fp16/hooks.py#L13
|
|
Version: v0.15.0
|
|
|
|
Models:
|
|
- Name: resnet50_b32x8_fp16_dynamic_imagenet
|
|
Metadata:
|
|
FLOPs: 4120000000
|
|
Parameters: 25560000
|
|
Epochs: 100
|
|
Batch Size: 256
|
|
Architecture:
|
|
- ResNet
|
|
In Collection: FP16
|
|
Results:
|
|
- Task: Image Classification
|
|
Dataset: ImageNet-1k
|
|
Metrics:
|
|
Top 1 Accuracy: 76.30
|
|
Top 5 Accuracy: 93.07
|
|
Weights: https://download.openmmlab.com/mmclassification/v0/fp16/resnet50_batch256_fp16_imagenet_20210320-b3964210.pth
|
|
Config: configs/fp16/resnet50_b32x8_fp16_dynamic_imagenet.py
|