import numpy as np
import torch
from model.trainer.fsl_trainer import FSLTrainer
from model.utils import (
pprint, set_gpu,
get_command_line_parser,
postprocess_args,
)
# from ipdb import launch_ipdb_on_exception
if __name__ == '__main__':
parser = get_command_line_parser()
args = postprocess_args(parser.parse_args())
# with launch_ipdb_on_exception():
pprint(vars(args))
set_gpu(args.gpu)
trainer = FSLTrainer(args)
trainer.train()
trainer.evaluate_test()
trainer.final_record()
print(args.save_path)