# Second Stage Test by Monash — Submission V3

## 1. Download the Docker Image

You can download the latest Docker image from [huggingface](https://huggingface.co/datasets/JianghaoWu/MBH-Seg_monash/blob/main/monash_v4_latest.tar.gz) as `monash_v4_latest.tar.gz`.

## 2. Load the Docker Image

To load the Docker image, use the following command:

`docker load -i monash_v4_latest.tar.gz`

After loading, you will see the message: `Loaded image: monash_v4:latest`, where `monash_v4:latest` is the name of the Docker image.


## 3. Test the Docker Image

To test the Docker container, run:

`docker container run --gpus "device=0" --name monash_v4 --rm --shm-size=24g -v $PWD/relative_path/test/fold/:/workspace/inputs/ -v $PWD/relative_path/outputs/fold/:/workspace/outputs/ monash_v4:latest /bin/bash -c "sh predict.sh"`


If you encounter any permission issues, you may need to run it with `sudo`:

`sudo docker container run --gpus "device=0" --name monash_v4 --rm --shm-size=24g -v $PWD/relative_path/test/fold/:/workspace/inputs/ -v $PWD/relative_path/outputs/fold/:/workspace/outputs/ monash_v4:latest /bin/bash -c "sh predict.sh"`


### Parameters:
- `--gpus "device=0"`: Select the appropriate GPU ID.
- Replace `relative_path/test/fold` with the **relative path** to your test folder, which should contain Nifti files (`.nii.gz`). This path is relative to your **current working directory**.
- Replace `relative_path/outputs/fold` with the **relative path** to your desired output folder, which will store the results. This path is also relative to your **current working directory**.
- `monash_v4:latest`: This is the Docker image name, which will be displayed after loading the image in step 2.
- `--shm-size=24g`: If you encounter memory issues, please increase the shared memory space, for example, --shm-size=32g.

or using absolute path:

`docker container run --gpus "device=0" --name monash_v4 --rm --shm-size=24g -v /absolute/path/to/your/test/fold/:/workspace/inputs/ -v /absolute/path/to/my/team/outputs/fold/:/workspace/outputs/ monash_v4:latest /bin/bash -c "sh predict.sh"`

- Replace `/absolute/path/to/your/test/fold/` with the **absolute path** to your test folder, which should contain Nifti files (`.nii.gz`).
- Replace `/absolute/path/to/my/team/outputs/fold/` with the **absolute path** to your desired output folder, where the output results will be stored.