提交 66d83fea 编写于 作者: R RachelXu7

Add ACT/FullQuant Demo

上级 bf4d6998
...@@ -33,15 +33,15 @@ TrainConfig: ...@@ -33,15 +33,15 @@ TrainConfig:
optimizer: optimizer:
type: Momentum type: Momentum
weight_decay: 0.000005 weight_decay: 0.000005
# origin_metric: 0.7532 origin_metric: 0.7532
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: /paddle/dataset/ILSVRC2012/ image_root: ./dataset/ILSVRC2012/
cls_label_path: /paddle/dataset/ILSVRC2012/train_list.txt cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
...@@ -69,8 +69,8 @@ DataLoader: ...@@ -69,8 +69,8 @@ DataLoader:
Eval: Eval:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: /paddle/dataset/ILSVRC2012/ image_root: ./dataset/ILSVRC2012/
cls_label_path: /paddle/dataset/ILSVRC2012/val_list.txt cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册