Awesome
pytorch-image-classification
use pytorch to do image classfiication tasks
使用方法
2020.05.06
此代码不在更新维护，请使用最新的 pytorch_img_classification_for_competition
2019.04.28
add test.py
version0.2.0
更新 手写了一份进度条工具，效果如下
loading train dataset
100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 4572/4572 [00:00<00:00, 1145746.42it/s]
loading train dataset
100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 2520/2520 [00:00<00:00, 1128994.45it/s]
Train Epoch: 1/40 [>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>] [Current: Loss 1.079473 Top1: 75.131233 ] 286/286 [ 100% ]
Val Epoch: 1/40 [>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>] [Current: Loss 0.469368 Top1: 89.484131 ] 79/79 [ 100% ]
Get Better top1 : 89.4841 saving weights to ./checkpoints/best_model/resnet50/0/model_best.pth.tar
Train Epoch: 2/40 [>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>] [Current: Loss 0.696026 Top1: 83.442696 ] 286/286 [ 100% ]
Val Epoch: 2/40 [>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>] [Current: Loss 0.275893 Top1: 91.269844 ] 79/79 [ 100% ]
Get Better top1 : 91.2698 saving weights to ./checkpoints/best_model/resnet50/0/model_best.pth.tar
Train Epoch: 3/40 [>>>>> ] [Current: Loss 0.667610 Top1: 84.165291 ] 31/286 [ 10% ]