run_train_base_fz3.sh

#!/bin/bash
CUDA_VISIBLE_DEVICES=0 python main.py \
    --data_dir=/workspace/EchoNet \
    --pretrained=pretrained/swin_base_patch4_window7_224_22k.pth \
    --batch_size=8 \
    --num_workers=4 \
    --accelerator=gpu \
    --variant=base \
    --max_epoch=20 \
    --frozen_stages 3