masashitsubaki/quantumdeepfield_molecule

Quantum deep field for molecule

quantumdeepfield_molecule's Issues

Generate the electron density map

Hello, the great work. I want to know how to generate the electron density map? Can you give me some help? thanks a lot.

Segmentation fault (core dumped) error

リポをclone後、デフォルト設定のまま $ bash train.sh したところ、
以下のエラーが出ました。
学習は走っているようですが、どうやらDataLoader周りでcrushしてしまっています。
ぜひこちらのコードを研究で活用させていただきたいのですが、手詰まりになってしまいました。
もし回避方法などあれば、ご教示いただけますと幸いです。

当方の実行環境は AWS (p2.xlarge) + Deep Learning AMI (Ubuntu 18.04) + RDKit です。
https://aws.amazon.com/marketplace/pp/B07Y43P7X5#pdp-overview

The code uses a GPU.
--------------------------------------------------
# of training samples:  34
# of validation samples:  4
# of test samples:  5
--------------------------------------------------
Set a QDF model.
# of model parameters: 392562
--------------------------------------------------
Start training of the QDF model with QM9under7atoms_atomizationenergy_eV dataset.
The training result is displayed in this terminal every epoch.
The result, prediction, and trained model are saved in the output directory.
Wait for a while...
The training will finish in about 0 hours 26 minutes.
--------------------------------------------------
Epoch	Time(sec)	Loss_E	Loss_V	MAE_val(eV)	MAE_test(eV)
0	0.8071885499994096	6209.085266113281	404.9622917175293	22.261384963989258	24.396270751953125
1	1.4956756869996752	5485.979797363281	392.7659606933594	20.992084503173828	23.13214683532715
2	2.175580770999659	4863.696319580078	386.9701862335205	19.829853057861328	21.962018966674805
3	2.8557513050000125	4676.126617431641	394.4252738952637	19.19409942626953	21.30068016052246
4	3.5467280840002786	4598.614196777344	386.27272605895996	20.13705062866211	22.27927589416504
5	4.246754251000311	5112.2559814453125	379.0389823913574	20.89983558654785	23.086027145385742
6	4.942768633000014	5517.936248779297	373.1512336730957	22.083404541015625	24.299663543701172
free(): invalid size
Traceback (most recent call last):
  File "/home/ubuntu/anaconda3/envs/torch_mi_apr21/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 986, in _try_get_data
    data = self._data_queue.get(timeout=timeout)
  File "/home/ubuntu/anaconda3/envs/torch_mi_apr21/lib/python3.7/queue.py", line 179, in get
    self.not_empty.wait(remaining)
  File "/home/ubuntu/anaconda3/envs/torch_mi_apr21/lib/python3.7/threading.py", line 300, in wait
    gotit = waiter.acquire(True, timeout)
  File "/home/ubuntu/anaconda3/envs/torch_mi_apr21/lib/python3.7/site-packages/torch/utils/data/_utils/signal_handling.py", line 66, in handler
    _error_if_any_worker_fails()
RuntimeError: DataLoader worker (pid 17362) is killed by signal: Aborted. 

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "train.py", line 389, in <module>
    MAE_val = tester.test(dataloader_val)[0]
  File "train.py", line 217, in test
    N = sum([len(data[0]) for data in dataloader])
  File "train.py", line 217, in <listcomp>
    N = sum([len(data[0]) for data in dataloader])
  File "/home/ubuntu/anaconda3/envs/torch_mi_apr21/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 517, in __next__
    data = self._next_data()
  File "/home/ubuntu/anaconda3/envs/torch_mi_apr21/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 1182, in _next_data
    idx, data = self._get_data()
  File "/home/ubuntu/anaconda3/envs/torch_mi_apr21/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 1138, in _get_data
    success, data = self._try_get_data()
  File "/home/ubuntu/anaconda3/envs/torch_mi_apr21/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 999, in _try_get_data
    raise RuntimeError('DataLoader worker (pid(s) {}) exited unexpectedly'.format(pids_str)) from e
RuntimeError: DataLoader worker (pid(s) 17362) exited unexpectedly
train.sh: line 39: 16905 Segmentation fault      (core dumped) python train.py $dataset $basis_set $radius $grid_interval $dim $layer_functional $hidden_HK $layer_HK $operation $batch_size $lr $lr_decay $step_size $iteration $setting $num_workers

question about the external potential

Dear authors, I have read your paper "Phys. Rev. Lett. 125, 206401" and enjoy it very much.

However I have a problem now about the choose of the external potential. In your paper your pick a gaussian potential as the target of HKS net. However, if I remember correctly, the true external potential corresponding to the electron density of a molecule is the coulomb potential of the each atom.

I just want to know whether it's true to use a coulomb potential instead of gaussian?

No questions now

No questions now