Skip to content

8mikehawk/scformer

BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

48 Commits
models
models
 
 
predict_results
predict_results
 
 
train_package
train_package
 
 
utils
utils
 
 
.gitignore
.gitignore
 
 
README.en.md
README.en.md
 
 
README.md
README.md
 
 
beta.py
beta.py
 
 
predict.py
predict.py
 
 
predict_polyp_final.py
predict_polyp_final.py
 
 
test.py
test.py
 
 
test_imageNet.py
test_imageNet.py
 
 
test_polyp.py
test_polyp.py
 
 
train.py
train.py
 
 
train_image_net.py
train_image_net.py
 
 
train_polyp.py
train_polyp.py
 
 
try_.py
try_.py
 
 

Repository files navigation

scformer (not finished)

usage

  1. in folder train_package, add your training folder for example sct_b1_isic2018.

  2. add test.conf and train.conf as the configurations. It looks like

    [dataset]
 name = ISIC2018
 train_img_root   = "your data path"
 val_img_root     = "your data path"

 train_label_root = "your data path"
 val_label_root   = "your data path"

 class_num = 2 

 crop_size = (512, 512)
 batch_size = 8
 num_workers = 8

 [model]
 # model config
 name = sct_b2

 [schedule]
 # training config
 device = cuda
 lr = 1e-4
 max_epoch = 10000
 checkpoint_save_path = /mnt/DATA-1/DATA-2/Feilong/scformer/train_package/sct_b1_isic2018/

 [logger]
 path = /mnt/DATA-1/DATA-2/Feilong/scformer/train_package/sct_b1_isic2018/sct_b1_isic.log

  3. add a new class on /utils/my_dataset.py if you want to train on a new dataset. Then, import this new class on the train.py like from utils.tools import ISIC2018

  5. 创建文件夹

  6. 改train.ocnf test.config

  7. 改train中数据集的类

  8. 确定train中model

  9. 改train中train_ds 那部分

About

No description, website, or topics provided.

Resources

Readme
Activity

Stars

0 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

 
 
 

Languages