README
0. Configuration File
The config file is ./configs/example.yaml
. You can modify test_file
or finetune
, which correspond to the CSV file location and the pre-trained model checkpoint location, respectively.
1. Model Weight Files and Dataset Configuration
Model Weight Files
The model weight files are located in the ./configs
folder. Ensure that these files are correctly loaded when running the model.
CSV File
The CSV file contains the paths to the test images and their corresponding labels. This file is used for testing the model.
Modifying test_example.py
Adjusting Chestxray14_Dataset
Class
In the Chestxray14_Dataset
class within test_example.py
, you need to adjust the column indices according to the structure of your CSV file. Specifically, ensure that the columns for image paths and labels are correctly referenced.
Updating valid_on
Method
The valid_on
method in test_example.py
contains a text_list
that should correspond to the labels in your CSV file. Make sure that the text_list
is updated to match the labels in your dataset.
2. Model Output
Prediction Results
The model will generate prediction results for each image and each class. These results will be saved as .npy
files.
Output Location
The .npy
files will be stored in subfolders within the ./results
directory. Each subfolder corresponds to a specific run or configuration of the model.