CAODH / EquiScore

MIT License
64 stars 6 forks source link

lmdb.Error: ./lmdbs/PDBscreen: No such file or directory #3

Open chentong456 opened 1 year ago

chentong456 commented 1 year ago
Process Process-1:
Traceback (most recent call last):
File "/home/chentong/anaconda3/envs/pytorch/lib/python3.9/multiprocessing/process.py", line 315, in _bootstrap
self.run()
File "/home/chentong/anaconda3/envs/pytorch/lib/python3.9/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/mnt/d/papaer_share/EquiScore/train.py", line 92, in run
train_dataset = ESDataset(train_keys,args, args.data_path,args.debug)#keys,args, data_dir,debug
File "/mnt/d/papaer_share/EquiScore/dataset/dataset.py", line 45, in __init__
env = lmdb.open(args.lmdb_cache, map_size=int(1e12), max_dbs=2, readonly=True)
lmdb.Error: ./lmdbs/PDBscreen: No such file or directory

我是按照Retraining EquiScore or fine tune your model is also very simnple!这个过程操作的。 先执行./data/data_splits/screen_model/data_split_for_training.py , 之后执行python train.py --ngpu 1 --train_keys your_keys_path --val_keys your_keys_path --test_keys your_keys_path 之后报上述错误,请问我如何得到the Folder containing trainging data in lmdb format?

CAODH commented 1 year ago

你可以检查一下文件路径问题,你是否已经通过邮件问过我相关问题,如果可以的话,请在这里提问,方便大家遇到相同问题可以参考。(You can check the file path problem, whether you have asked me about it through email, if possible, please ask here, so that everyone can refer to the same problem.)

chentong456 commented 1 year ago

就是没有这个路径,但是我不知道这个路径里面的数据是什么,以及如何创建

CAODH commented 1 year ago

你打开test参数 ,python train.py --ngpu 1 --test --train_keys your_keys_path --val_keys your_keys_path --test_keys your_keys_path ,应该就行了,不过这样会很慢,推荐你根据readme提供的方案,先预处理数据集到LMDB 请参考readme的命令,推荐你使用bash_script中的命令试一下,有任何问题都可以及时交流 祝好!

Mengo-ye commented 4 months ago

你好,想请教下如何准备这个文件./lmdbs/PDBscreen,谢谢。