装火把检查点的问题--存储错误，大小错误--如何修复？

Question

我得到了以下错误：

Traceback (most recent call last):
  File "/home/miranda9/diversity-for-predictive-success-of-meta-learning/div_src/diversity_src/experiment_mains/main2_distance_sl_vs_maml.py", line 790, in <module>
    main_data_analyis()
  File "/home/miranda9/diversity-for-predictive-success-of-meta-learning/div_src/diversity_src/experiment_mains/main2_distance_sl_vs_maml.py", line 597, in main_data_analyis
    args.mdl2 = get_sl_learner(args)
  File "/raid/projects/miranda9/diversity-for-predictive-success-of-meta-learning/div_src/diversity_src/data_analysis/common.py", line 195, in get_sl_learner
    model = load_original_rfs_ckpt(args, path_to_checkpoint=args.path_2_init_sl)
  File "/raid/projects/miranda9/diversity-for-predictive-success-of-meta-learning/div_src/diversity_src/data_analysis/common.py", line 168, in load_original_rfs_ckpt
    ckpt = torch.load(path_to_checkpoint, map_location=args.device)
  File "/home/miranda9/miniconda3/envs/meta_learning_a100/lib/python3.9/site-packages/torch/serialization.py", line 608, in load
    return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
  File "/home/miranda9/miniconda3/envs/meta_learning_a100/lib/python3.9/site-packages/torch/serialization.py", line 794, in _legacy_load
    deserialized_objects[key]._set_from_file(f, offset, f_should_read_directly)
RuntimeError: storage has wrong size: expected 0 got 64

为什么和如何修复它？

相关：https://discuss.pytorch.org/t/runtimeerror-storage-has-wrong-size/88109/4

Answer 1

通过简单地将模型ckpt文件重新上传到集群中，这个问题就从我这里消除了。

其他观点见：https://discuss.pytorch.org/t/runtimeerror-storage-has-wrong-size/88109/4