Welcome to DeepModeling Tutorial!

Hi everyone, here are the tutorials for DeepModeling Projects.

Using DeePMD-Kit

This tutorial tell you how to use DeePMD-kit，for detail information, you can check DeePMD-kit Doc

Introduction

Over the last few decades, molecular dynamics (MD) simulations have attracted much attention due to their wide range of applications in many fields such as condensed matter physics, materials science, polymer chemistry, and molecular biology. They provide researchers access to examine the behavior of atoms or molecules, which is valuable and has the potential to enrich our knowledge, especially when experimentation is difficult, expensive, or even impossible.

It is well recognized that the quality of MD simulations is ultimately limited by the accuracy of the PES and accurately representing the PES is an important challenge in the field of MD simulations. The empirical atomic potential models and the quantum mechanical models have long been two types of models commonly used. The empirical atomic potential models consist of simple low-dimensional terms. They often show excellent computational efficiency, but have limited accuracy. The quantum mechanical models determine the energies and forces on atoms by approximately solving the Schrödinger equation for electronic structure and exhibit higher accuracy. However, quantum mechanical models are computationally demanding, and it is not quite practical for either large-scale or long-time calculations. Overall, a dilemma exists between the choice of an empirical atomic potential model for high efficiency and that of a quantum mechanics model for high accuracy.

Recently, machine-learning (ML) models are emerging as useful tools to address this dilemma. Descriptors and ML algorithms are two main components of the current ML models. The former is used to guarantee the natural symmetries of the system and the latter is used to establish a direct functional relationship between atomic configurations and potential energy by training on reference data generated by quantum mechanics. Once trained, ML models can provide the same accuracy as the quantum mechanical method used to generate the reference data, such as the density functional theory (DFT) accuracy. Meanwhile, the computational cost of DFT scales cubic and that of ML models scale linearly with the size of the system.

So far, different types of ML models have been reported in literatures, such as Behler-Parrinello neural network potentials (BPNNP)[1], Gaussian approximation potentials (GAP)[2], spectral neighbor analysis potentials (SNAP)[3], ANI-1[4], SchNet[5] and Deep Potentials (DP) [6,7,8]. It is worth pointing out that, despite great successes, there are still many challenging issues remaining to be tackled[9]. For example, the neglect of interactions beyond the cut-off radius may lead to systematic prediction errors[10].

This chapter focuses on the DP models. In addition to enabling quantum mechanical accuracy, current DP models also have the following characteristics: (i) ease to preserve the symmetry of the system, especially when there are multiple elemental species; (ii) high computational efficiency, being at least five orders of magnitude faster than DFT; (iii) the model is end-to-end and therefore has little human intervention; (iv) support for MPI and GPU, making it highly efficient on modern heterogeneous high performance supercomputers. Thanks to this, the DP models have been successfully employed in studies of water and water-containing systems[11,12,13,14], metals and alloys[15,16,17,18], phase diagrams[19,20,21], high-entropy ceramics[22,23], chemical reaction[24,25,26], solid-state electrolytes[27], ionic liquids[28], etc. We refer to Ref.[29] for a recent review of DP for materials systems.

Installation

There are three easy methods to install DeePMD-kit：

Install off-line packages
Install with conda
Install with docker

Users can choose a suitable method depending on the machine environment. The following is a detailed description of these three methods.

Install off-line packages

Users can use the offline packages to install the DeePMD-kit if their machine cannot be connected to the internet. Both CPU and GPU version offline packages are available on the Releases page.

Some GPU version off-line packages are splited into two files due to size limit of GitHub. Users can merge them into one with the cat command and then install the DeePMD-kit software with the bash command.

$ cat deepmd-kit-2.0.0-cuda11.3_gpu-Linux-x86_64.sh.0 deepmd-kit-2.0.0-cuda11.3_gpu-Linux-x86_64.sh.1 > deepmd-kit-2.0.0-cuda11.3_gpu-Linux-x86_64.sh
$ bash deepmd-kit-2.0.0-cuda11.3_gpu-Linux-x86_64.sh

During the installation, users need to specify the installation path of the DeePMD-kit. It is assumed that the user chose to install DeePMD-kit at “/root/deepmd-kit”. Then, users need to configure the environment variable.

$ export PATH="/root/deepmd-kit/bin/:$PATH"

Users should remember to configure the environment variable after opening a new terminal. Users can also add the above line into the bashrc file, which is not explained here.

Install with conda

Users can use conda to install the DeePMD-kit if their machine can be connected to the internet. Before installing DeePMD-kit, users need to install Anaconda or Miniconda and activate the conda environment.

Both the CPU and GPU versions of DeePMD-kit can be installed via conda. Users can create an environment that contains the CPU version of DeePMD-kit and LAMMPS.

(base)$ conda create -n deepmd deepmd-kit=*=*cpu libdeepmd=*=*cpu lammps -c https://conda.deepmodeling.org

or create an environment that contains the GPU version of DeePMD-kit and LAMMPS.

(base)$ conda create -n deepmd deepmd-kit=*=*gpu libdeepmd=*=*gpu lammps cudatoolkit=11.3 horovod -c https://conda.deepmodeling.org

The environment also contains the CUDA Toolkit. Users could change the CUDA Toolkit version from 10.1 or 11.3. The latest version of DeePMD-kit will be installed by the above command. Users may want to specify the DeePMD-kit version such as 2.0.0 using

(base)$ conda create -n deepmd deepmd-kit=2.0.0=*cpu libdeepmd=2.0.0=*cpu lammps horovod -c https://conda.deepmodeling.org

Before using the DeePMD-kit, users need to ensure that the environment is active. Users can enable the deepmd environment using

(base)$ conda activate deepmd

When the environment is activated, (base) will be converted to (deepmd) on the left in the terminal. For example,

(deepmd)$

Install with docker

Users can also use docker to install the DeePMD-kit if their machine can be connected to the internet.

To pull the CPU version:

$ docker pull ghcr.io/deepmodeling/deepmd-kit:2.0.0_cpu

To pull the GPU version:

$ docker pull ghcr.io/deepmodeling/deepmd-kit:2.0.0_cuda10.1_gpu

To pull the ROCm version:

$ docker pull deepmodeling/dpmdkit-rocm:dp2.0.3-rocm4.5.2-tf2.6-lmp29Sep2021

Verify the installation

If the installation is successful, DeePMD-kit (dp) and LAMMPS (lmp) will be available to execute.mpirun is also available considering users may want to train models or run LAMMPS in parallel. To verify the installation, users can execute

$ dp -h

the terminal will show the help information like

usage: dp [-h] [--version]{config,transfer,train,freeze,test,compress,doc-traininput,model-devi,convert-from}

DeePMD-kit: A deep learning package for many-body potential energy representation and molecular dynamics
...

Note that users can also install the DeePMD-kit software from the source code, but this process is relatively complex. A detailed description is presented in ‘Install from source code’ Section of DeePMD-kit’s documentation.

Theory

Before introducing the DP method, we define the coordinate matrix $\boldsymbol{\mathcal{R}} \in \mathbb{R}^{N \times 3}$ of a system containing $N$ atoms,

\[\mathcal{R}=\left\{\boldsymbol{r}_{1}^{T}, \cdots, \boldsymbol{r}_{i}^{T}, \cdots, \boldsymbol{r}_{N}^{T}\right\}^{T}, \boldsymbol{r}_{i}=\left(x_{i}, y_{i}, z_{i}\right),(1)\]

$\boldsymbol{r}_{i}$ contains 3 Cartesian coordinates of atom $i$ and $\boldsymbol{\mathcal{R}}$ can be transformed into local environment matrices $\left\{\boldsymbol{\mathcal{R}}^{i}\right\}_{i=1}^{N}$,

\[\boldsymbol{\mathcal{R}}^{i}=\left\{\boldsymbol{r}_{1 i}^{T}, \cdots, \boldsymbol{r}_{j i}^{T}, \cdots, \boldsymbol{r}_{N_{i}, i}^{T}\right\}^{T}, \boldsymbol{r}_{j i}=\left(x_{j i}, y_{j i}, z_{j i}\right),(2)\]

where $j$ and $N_{i}$ are indexes of neighbors of atom $i$ within the cut-off radius $r_{c}$, and $\boldsymbol{r}_{j i} \equiv \boldsymbol{r}_{j}-\boldsymbol{r}_{i}$ is defined as the relative coordinate.

In the DP method, the total energy $E$ of a system is constructed as a sum of atomic energies.

\[ E=\sum_{i} E_{i},(3)\]

with $E_{i}$ being the local atomic energy of the atom $i$. $E_{i}$ depends on the local environment of the atom $i$:

\[ E=\sum_{i} E_{i}=\sum_{i} E\left(\mathcal{R}^{i}\right),(4)\]

The mapping of $\boldsymbol{\mathcal{R}}^{i}$ to $E_{i}$ is constructed in two steps. As seen in , $\boldsymbol{\mathcal{R}}^{i}$ is first mapped to a feature matrix, also called the descriptor, $\boldsymbol{\mathcal{D}}^{i}$ to preserve the translational, rotational, and permutational symmetries of the system. $\boldsymbol{\mathcal{R}}^{i} \in \mathbb{R}^{N_{i} \times 3}$ is first transformed into generalized coordinate $\tilde{\boldsymbol{\mathcal{R}}}^{i} \in \mathbb{R}^{N_{i} \times 4}$ .

\[ \left\{x_{j i}, y_{j i}, z_{j i}\right\} \mapsto\left\{s\left(r_{j i}\right), \hat{x}_{j i}, \hat{y}_{j i}, \hat{z}_{j i}\right\},(5)\]

where $\hat{x}_{j i}=\frac{s\left(r_{j i}\right) x_{j i}}{r_{j i}}$, $\hat{y}_{j i}=\frac{s\left(r_{j i}\right) y_{j i}}{r_{j i}}$, and $\hat{z}_{j i}=\frac{s\left(r_{j i}\right) z_{j i}}{r_{j i}}$. $s\left(r_{j i}\right)$ is a weighting function to reduce the weight of particles that are more distant from the atom $i$, defined as:

\[\begin{split}s\left(r_{j i}\right)= \begin{cases}\frac{1}{r_{j i}}, & r_{j i}<r_{c s} \\ \frac{1}{r_{j i}} \{ {(\frac{r_{j i} - r_{c s}}{ r_c - r_{c s}})}^3 (-6 {(\frac{r_{j i} - r_{c s}}{ r_c - r_{c s}})}^2 +15 \frac{r_{j i} - r_{c s}}{ r_c - r_{c s}} -10) +1 \}, & r_{c s}<r_{j i}<r_{c} \\ 0, & r_{j i}>r_{c}\end{cases},(6)\end{split}\]

here $r_{j i}$ is the Euclidean distance between atoms $i$ and $j$, and $r_{cs}$ is the smooth cutoff parameter. By introducing $s\left(r_{j i}\right)$ the components in $\tilde{\boldsymbol{\mathcal{R}}}^{i}$ smoothly go to zero from $r_{cs}$ to $r_{c}$. Then $s\left(r_{j i}\right)$, i.e. the first column of $\tilde{\boldsymbol{\mathcal{R}}}^{i}$, is mapped to a embedding matrix $\mathcal{G}^{i 2} \in \mathbb{R}^{N_{i} \times M_{1}}$, through an embedding neural network. By taking the first $M_{2}(<M_{1})$ columns of $\boldsymbol{\mathcal{G}}^{i 2} \in \mathbb{R}^{N_{i} \times M_{2}}$, we obtain another embedding matrix $\mathcal{G}^{i 2} \in \mathbb{R}^{N_{i} \times M_{2}}$. Finally, we define the feature matrix $\mathcal{G}^{i 2} \in \mathbb{R}^{M_{1} \times M_{2}}$ of atom $i$:

\[\mathcal{D}^{i}=\left(\mathcal{G}^{i 1}\right)^{T} \tilde{\mathcal{R}}^{i}\left(\tilde{\mathcal{R}}^{i}\right)^{T} \mathcal{G}^{i 2},(7)\]

In feature_matrix, translational and rotational symmetries are preserved by the matrix product of $\tilde{\mathcal{R}}^{i}\left(\tilde{\mathcal{R}}^{i}\right)^{T}$, and permutational symmetry is preserved by the matrix product of $\left(\mathcal{G}^{i}\right)^{T} \tilde{\mathcal{R}}^{i}$. Next, each $\boldsymbol{\mathcal{D}}^{i}$ is mapped to a local atomic energy $E_{i}$ through a fitting network.

Both the embedding network $\mathcal{N}^e$ and fitting network $\mathcal{N}^f$ are feed-forward neural networks containing several hidden layers. The mapping from input data $d_{l}^{\mathrm{in}}$ of the previous layer to output data $d_{k}^{\mathrm{out}}$ of the next layeris composed of a linear and a non-linear transformation.

\[d_{k}^{o u t}=\varphi\left(\sum_{k l} w_{k l} d_{l}^{i n}+ b_{k}\right),(8)\]

In Eq.(8), ${w}_{k l}$ is the connecting weight, ${b}_{k}$ the bias weight, and $\varphi$ is a non-linear activation function. It needs to be noted that only linear transformations are applied at the output nodes. The parameters contained in the embedding and fitting networks are obtained by minimizing the loss function $L$:

\[L\left(p_{\epsilon}, p_{f}, p_{\xi}\right)=\frac{p_{\epsilon}}{N} \Delta \epsilon^{2}+\frac{p_{f}}{3 N} \sum_{i}\left|\Delta \boldsymbol{F}_{i}\right|^{2}+\frac{p_{\xi}}{9N}\|\Delta \xi\|^{2},(9)\]

where $\Delta \epsilon$, $\Delta \boldsymbol{F}_{i}$, and $\Delta \xi$ denote root mean square error (RMSE) in energy, force, and virial, respectively. During the training process, the prefactors $p_{\epsilon}$, $p_{f}$, and $p_{\xi}$ are determined by

\[p(t)=p^{\operatorname{limit}}\left[1-\frac{r_{l}(t)}{r_{l}^{0}}\right]+p^{\operatorname{start}}\left[\frac{r_{l}(t)}{r_{l}^{0}}\right],(10)\]

where $r_{l}(t)$ and $r_{l}^{0}$ are the learning rate at training step $t$ and training step 0. $r_{l}(t)$ is defined as

\[r_{l}(t)=r_{l}^{0} \times d_{r}^{t / d_{s}},(11)\]

where $d_{r}$ and $d_{s}$ are the decay rate and decay steps, respectively. The decay rate $d_{r}$ is required to be less than 1. The reader is referred to the original papers of DeepPot-SE (DP) method for details.

How to Setup a DeePMD-kit Training within 5 Minutes

DeePMD-kit is a software to implement Deep Potential. There is a lot of information on the Internet, but there are not so many tutorials for the new hand, and the official guide is too long. Today, I’ll take you 5 minutes to get started with DeePMD-kit.

Let’s take a look at the training process of DeePMD-kit:

Prepare data –> Training –> Freeze/Compress the model

What? Only three steps? Yes, it’s that simple.

Preparing data is converting the computational results of DFT to data that can be recognized by the DeePMD-kit.
Training is train a Deep Potential model using the DeePMD-kit with data prepared in the previous step.
Finally, what we need to do is to freeze/compress the restart file in the training process into a model. I believe you can’t wait to get started. Let’s go!

Tutorial Data

First, let’s download and decompress the tutorial data:

    $ wget https://dp-public.oss-cn-beijing.aliyuncs.com/community/DeePMD-kit-FastLearn.tar
    $ tar xvf DeePMD-kit-FastLearn.tar

Then we can go to the Tutorial data and have a look:

    $ cd DeePMD-kit-FastLearn
    $ ls
    00.data 01.train data

Three directories are set for different purpose:

00.data: contains an example of VASP result OUTCAR
01.train: contains an example of DeePMD-kit configuration input.json
data: contains an example of DeePMD-kit training/validation data

Preparing Data

Now Let’s go into 00.data directory:

    $ cd 00.data
    $ ls
    OUTCAR

This file OUTCAR is the computational result of the VASP. We need convert it into DeePMD-kit format. The data format of the DeePMD-kit is introduced in the official document but seems complex. Don’t worry, I’d like to introduce a data processing tool: dpdata! You can use only one line Python scripts to process data. So easy!

   import dpdata
   dpdata.LabeledSystem('OUTCAR').to('deepmd/npy', 'data', set_size=200)

In this example, we converted the computational results of the VASP in the OUTCAR to the data format of the DeePMD-kit and saved in to a directory named data, where npy is the compressed format of the numpy, which is required by the DeePMD-kit training.

Suppose you have an “OUTCAR” for molecular dynamics, which contains 1000 frames.set_size=200 means these 1000 points will be divided into 5 subsets, which is named as data/set.000~data/set.004, respectively. The size of each set is 200. In these 5 sets, data/set.000~data/set.003 will be considered as the training set by the DeePMD-kit, and data/set.004 will be considered as the test set. The last set will be considered as the test set by the DeePMD-kit by default. If there is only one set, the set will be both the training set and the test set. (Of course, such test set is meaningless.)

“OUTCAR” we provided only contains 1 frame, so in “data” directory(in the same directory with “OUTCAR”) there is only 1 set: data/set.000. Some procudure needs to be done if you want to use these data. Detailed method using dpdata can be found in next chapter.

Now we just skip these details and use the data we prepared for you. The data is in the root directory of our Tutorial data.

    $ cd ..
    $ ls
    00.data 01.train data

Training

It’s required to prepare an input script to start the DeePMD-kit training. Are you still out of the fear of being dominated by INCAR script? Don’t worry, it’s much easier to configure the DeePMD-kit than configuring the VASP.We have prepared input.json for you, you can find it in “01.train” directory:

   $ cd 01.train
   $ ls
   input.json

The strength of the DeePMD-kit is that the same training parameters are suitable for different systems, so we only need to slightly modify input.json to start training. Here is the first parameter to modify:

   "type_map":     ["O", "H"],

In the DeePMD-kit data, each atom type is numbered as an integer starting from 0. The parameter gives an element name to each atom in the numbering system. Here, we can copy from the content of data/type_map.raw. For example,

   "type_map":    ["A", "B","C"],

Next, we are going to modify the neighbour searching parameter:

   "sel":       [46, 92],

Each number in this list gives the maximum number of atoms of each type among neighbor atoms of an atom. For example, 46 means there are at most 46 O (type 0) neighbours. Here, our elements were modified to A, B, and C, so this parameters is also required to modify. What to do if you don’t know the maximum number of neighbors? You can be roughly estimate one by the density of the system, or try a number blindly. If it is not big enough, the DeePMD-kit will shoot WARNINGS. Below we changed it to

   "sel":       [64, 64, 64]

In addtion, we need to modify "systems" in "training_data"

   "training_data":{
          "systems":     ["../data/data_0/", "../data/data_1/", "../data/data_2/"],

and "validation_data"

   "validation_data":{
          "systems":     ["../data/data_3"],

Here I’d like to introduce the definition of the data system. The DeePMD-kit considers that data with corresponding element types and atomic numbers form a system. If data cannot be put into a system, multiple systems is required to be set as a list here:

   "systems": ["system1", "system2"]

Finnally, we are likely to modify another parameter:

   "numb_steps":   1000,

numb_steps is the numebr of training step using the SGD method of deep learning.(It is only an example, you should set a larger number in practice)

Now we have succesfully set a input file! To start training, we execuate

   dp train input.json

and wait for results. During the training process, we can see lcurve.out to observe the error reduction.Among them, Column 4 and 5 are the test and training errors of energy (normalized by the number of atoms), and Column 6 and 7 are the test and training errors of the force.

Freeze/Compress the Model

After training, we can use the following script to freeze the model:

   dp freeze -o graph.pb

The default filename of the output model is graph.pb. And we can use the following script to compress the model:

   dp compress -i graph.pb -o graph-compress.pb

As so, we have got a good or bad DP model. As for the reliability of this model and how to use it, I will give you a detailed tutorial in the next post.

Handson-Tutorial(v2.0.3)

Workflow of the DeePMD-kit

The workflow of the DeePMD-kit contains three parts:

Data preparation: Training data is generated based on ab-initio calculations and format conversion is performed;
Model training: Prepare input script and then train the DP model using the data and script;
Model application: Use the obtained DP model for MD simulation or model inference.

Example: a gas-phase methane molecule

The following introduces the basic usage of the DeePMD-kit, taking a gas-phase methane molecule as an example.

Data preparation

Preparing the training data includes both generating ab-initio training data and converting the data format, which is the first step of training a DP model with DeePMD-kit.

Training data is often generated using ab-initio molecular dynamics (AIMD) simulations and needs to be converted to a format that can be used directly by DeePMD-kit.

The files needed for this tutorial are available.

$ wget https://dp-public.oss-cn-beijing.aliyuncs.com/community/CH4.tar
$ tar xvf CH4.tar

Go to and check the CH4 folder：

$ cd CH4
$ ls
00.data 01.train 02.lmp

There are 3 folders here:

The folder 00.data contains the data
The folder 01.train contains an example input script to train a model with DeePMD-kit
The folder 02.lmp contains the LAMMPS example script for molecular dynamics simulation

AIMD data generation

The training data of the DeePMD-kit contains the atom type, the simulation box, the atom coordinate, the atom force, the system energy, and the virial. A snapshot of a molecular system that has this information is called a frame. A system of data includes many frames that share the same number of atoms and atom types. For example, a molecular dynamics trajectory can be converted into a system of data, with each time step corresponding to a frame in the system.

As this tutorial is about the DeePMD-kit, training data generated by AIMD simulations is provided.

Go to and check the 00.data folder

$ cd 00.data
$ ls 
OUTCAR

The OUTCAR was produced by an AIMD simulation of a gas-phase methane molecule using VASP.

Data format conversion

The DeePMD-kit adopts a compressed data format. All training data should first be converted into this format and can then be used by DeePMD-kit. The data format is explained in detail in the DeePMD-kit manual that can be found in the ‘data’ Section of DeePMD-kit’s documentation.

We provide a convenient tool named dpdata for converting the data produced by VASP, Gaussian, Quantum-Espresso, ABACUS, and LAMMPS into the compressed format of DeePMD-kit. For details about dpdata, see dpdata’s documentation.

Users can install dpdata via

$ git clone https://github.com/deepmodeling/dpdata.git dpdata  
$ cd dpdata  
$ python setup.py install

or

$ pip install dpdata

Data format conversion using dpdata can be completed in two steps: load data and dump data. Now start an interactive python environment, for example

$ python

then execute the following commands:

import dpdata 
import numpy as np
data = dpdata.LabeledSystem('OUTCAR', fmt = 'vasp/outcar') 
print('# the data contains %d frames' % len(data))

On the screen, you can see that the OUTCAR file contains 200 frames of data. We randomly pick 40 frames as validation data and the rest as training data.

# random choose 40 index for validation_data
index_validation = np.random.choice(200,size=40,replace=False)     
# other indexes are training_data
index_training = list(set(range(200))-set(index_validation))       
data_training = data.sub_system(index_training)
data_validation = data.sub_system(index_validation)
# all training data put into directory:"training_data" 
data_training.to_deepmd_npy('training_data')               
# all validation data put into directory:"validation_data"
data_validation.to_deepmd_npy('validation_data')           
print('# the training data contains %d frames' % len(data_training)) 
print('# the validation data contains %d frames' % len(data_validation)) 

The commands import a system of data from the OUTCAR (with format vasp/outcar), and then dump it into the compressed format (numpy compressed arrays).

Now users have completed the data conversion. The data in DeePMD-kit format is stored in the folder 00.data. Let’s have a look:

$ ls 
OUTCAR training_data validation_data

The directories “training_data” and “validation_data” have a similar structure, so we just explain “training_data”:

$ ls training_data
set.000 type.raw type_map.raw

set.000 is a directory, containing data in compressed format (numpy compressed arrays).
type.raw is a file, containing types of atoms(Represented in integer)
type_map.raw is a file, containing the type name of atoms.

Lets have a look at type.raw:

$ cat training_data/type.raw 
0 0 0 0 1

This tells us there are 5 atoms in this example, 4 atoms represented by type “0”, and 1 atom represented by type “1”. Sometimes users needs to map the integer types to atom name. The mapping can be given by the file type_map.raw

$ cat training_data/type_map.raw 
H C

This tells us the type “0” is named by “H”, and the type “1” is named by “C”.

input script

Once the data preparation is done, we can go on with training. Now go to the training directory

$ cd ../01.train
$ ls 
input.json

where input.json gives you an example training script. Users can specify the training process by specifying the value of keywords in input.json. The keywords are explained in detail in the DeePMD-kit manual, so they are not comprehensively explained here.

The keywords in input.json can be divided into 4 sections

Model: define the descriptor that maps atomic configuration to a set of symmetry invariant features, and the fitting net that takes descriptor as input and predicts the atomic contribution to the target physical property;
Learning rate: define the start learning rate, stop learning rate, decays steps, etc.
Loss function: define the type of loss, prefactor of energy, force and virial, etc.
Training: define the path of the training dataset and validation dataset, training steps, etc.

Model

The model keywords are given in the following:

"model":{
    "type_map":    ["H", "C"],                           
    "descriptor":{
        "type":            "se_e2_a",                    
        "rcut":            6.00,                        
        "rcut_smth":       0.50,                         
        "sel":             [4, 1],                       
        "neuron":          [10, 20, 40],                 
        "resnet_dt":       false,
        "axis_neuron":     4,                            
        "seed":            1,
        "_comment":        "that's all"
    },
    "fitting_net":{
        "neuron":          [100, 100, 100],   
        "resnet_dt":       true,
        "seed":            1,
        "_comment":        "that's all"
    },
    "_comment":    "that's all"
},

Description of keywords:

keywords	type	Description
type_map	list	Give the name to each type of atoms.
descriptor	dict	The descriptor of atomic environment.
type	str	The type of the descritpor.
sel	list	sel_a[i] gives the selected number of type-i neighbors.
rcut	float	The cut-off radius.
rcut_smth	float	Where to start smoothing. For example the 1/r term is smoothed from rcut to rcut_smth
neuron	list	Number of neurons in each hidden layers of embedding net.
axis_neuron	int	Size of the submatrix of G (embedding matrix)
seed	int	Random seed for parameter initialization.
fitting_net	dict	The fitting of physical properties.
neuron	list	Number of neurons in each hidden layers of fitting net.

Description of example: The se_e2_a descriptor is used to train the DP model. The cut-off radius is set to 6 Å and the components in $\tilde{\mathcal{R}}^{i}$ smoothly go to zero from 0.5 to 6 Å. Within the cut-off radius, the local environment of the H-atom is determined by 4 nearest-neighbour, and the local environment of the C-atom is determined by 1 nearest neighbour atom. The size of the embedding and fitting network to [10, 20, 40] and [100, 100, 100], respectively.

Learning rate

The learning_rate keywords are given in the following:

"learning_rate" :{
    "type":                "exp",
    "decay_steps":         5000,
    "start_lr":            0.001,    
    "stop_lr":             3.51e-8,
    "_comment":            "that's all"
},

Description of keywords:

keywords	type	Description
learning_rate	dict	The definition of learning rate.
type	str	The type of the learning rate.
decay_steps	int	The learning rate is decaying every this number of training steps.
start_lr	float	The learning rate the start of the training.
stop_lr	float	The desired learning rate at the end of the training.

Description of example:

During the training, the learning rate decays exponentially from start_lr to stop_lr The starting learning rate, stop learning rate, and decay steps are set to 0.001, 3.51e-8, and 5000, respectively.

Loss

The loss keywords are given in the following:

"loss" :{
    "type":                "ener",
    "start_pref_e":        0.02,
    "limit_pref_e":        1,
    "start_pref_f":        1000,
    "limit_pref_f":        1,
    "start_pref_v":        0,
    "limit_pref_v":        0,
    "_comment":            "that's all"
},

Description of keywords:

keywords	type	Description
loss	dict	The definition of loss function.
type	str	The type of the loss.
start_pref_e	float	The prefactor of energy loss at the start of the training.
limit_pref_e	float	The prefactor of energy loss at the limit of the training.
start_pref_f	float	The prefactor of force loss at the start of the training.
limit_pref_f	float	The prefactor of force loss at the limit of the training.
start_pref_v	float	The prefactor of virial loss at the start of the training.
limit_pref_v	float	The prefactor of virial loss at the limit of the training.

Description of example:

The loss function of the DeePMD-kit is determined by weighting the force, energy, and virial. In the loss function, pref_e increases from 0.02 to 1, and pref_f decreases from 1000 to 1 progressively, which means that the force term dominates at the beginning, while energy and virial terms become important at the end. This strategy is very effective and reduces the total training time. pref_v is set to 0, indicating that no virial data are included in the training process.

Training

The training keywords are given in the following

"training" : {
	"training_data": {
		"systems":            ["../00.data/training_data"],     
		"batch_size":         "auto",                       
		"_comment":           "that's all"
	},
	"validation_data":{
		"systems":            ["../00.data/validation_data/"],
		"batch_size":         "auto",               
		"numb_btch":          1,
		"_comment":           "that's all"
	},
	"numb_steps":             100000,                           
	"seed":                   10,
	"disp_file":              "lcurve.out",
	"disp_freq":              1000,
	"save_freq":              10000,
}

Description of keywords:

keywords	type	Description
training	dict	The definition of training.
training_data	dict	Configurations of training data.
systems	str	The data systems for training.
batch_size	list, str, or int	str “auto”: automatically determines the batch size so that the batch_size times the number of atoms in the system is no less than 32.
validation_data	dict	Configurations of validation data.
numb_btch	int	An integer that specifies the number of systems to be sampled for each validation period.
numb_steps	int	Number of training batch. Each training uses one batch of data.
disp_file	str	The file for printing learning curve.
disp_freq	int	The frequency of printing learning curve.
save_freq	int	The frequency of saving check point

Description of example:

During the training, the training data is at “../00.data/validation_data/”, and validation data is at “../00.data/validation_data/”. The model is trained for $10^6$ steps. The learning curve is written to the lcurve.out every 1000 steps, and the model-related files are saved every 10000 steps.

Train process

The following describes the training process of the DP model using the DeePMD-kit.

Start training Users can start the training with DeePMD-kit by simply running

$ dp train input.json

On the screen, you see the information of the data system(s)

DEEPMD INFO      ----------------------------------------------------------------------------------------------------
DEEPMD INFO      ---Summary of DataSystem: training     -------------------------------------------------------------
DEEPMD INFO      found 1 system(s):
DEEPMD INFO                              system        natoms        bch_sz        n_bch          prob        pbc
DEEPMD INFO           ../00.data/training_data/             5             7           22         1.000          T
DEEPMD INFO      -----------------------------------------------------------------------------------------------------
DEEPMD INFO      ---Summary of DataSystem: validation   --------------------------------------------------------------
DEEPMD INFO      found 1 system(s):
DEEPMD INFO                               system       natoms        bch_sz        n_bch          prob        pbc
DEEPMD INFO          ../00.data/validation_data/            5             7            5         1.000          T

and the starting and final learning rate of this training

DEEPMD INFO      start training at lr 1.00e-03 (== 1.00e-03), decay_step 5000, decay_rate 0.950006, final lr will be 3.51e-08

If everything works fine, you will see, on the screen, information printed every 1000 steps, like

DEEPMD INFO    batch    1000 training time 7.61 s, testing time 0.01 s
DEEPMD INFO    batch    2000 training time 6.46 s, testing time 0.01 s
DEEPMD INFO    batch    3000 training time 6.50 s, testing time 0.01 s
DEEPMD INFO    batch    4000 training time 6.44 s, testing time 0.01 s
DEEPMD INFO    batch    5000 training time 6.49 s, testing time 0.01 s
DEEPMD INFO    batch    6000 training time 6.46 s, testing time 0.01 s
DEEPMD INFO    batch    7000 training time 6.24 s, testing time 0.01 s
DEEPMD INFO    batch    8000 training time 6.39 s, testing time 0.01 s
DEEPMD INFO    batch    9000 training time 6.72 s, testing time 0.01 s
DEEPMD INFO    batch   10000 training time 6.41 s, testing time 0.01 s
DEEPMD INFO    saved checkpoint model.ckpt

They present the training and testing time counts. At the end of the 10000th batch, the model is saved in Tensorflow’s checkpoint file model.ckpt. At the same time, the training and testing errors are presented in file lcurve.out.

Users can check the lcurve.out using the cat command after training

$ cat lcurve.out
#step       rmse_val       rmse_trn       rmse_e_val       rmse_e_trn       rmse_f_val       rmse_f_trn           lr
      0     1.34e+01       1.47e+01         7.05e-01         7.05e-01         4.22e-01         4.65e-01     1.00e-03
    ...
 999000     1.24e-01       1.12e-01         5.93e-04         8.15e-04         1.22e-01         1.10e-01      3.7e-08
1000000     1.31e-01       1.04e-01         3.52e-04         7.74e-04         1.29e-01         1.02e-01      3.5e-08

The lcurve.out contains 8 columns, from left to right, which are the training step, the validation loss, training loss, root mean square (RMS) validation error of energy, RMS training error of energy, RMS validation error of force, RMS training error of force and the learning rate. The RMS error (RMSE) of the energy is normalized by the number of atoms in the system. It is demonstrated that after $10^6$ steps of training, the energy testing error is less than 1 meV and the force testing error is around 120 meV/Å. It is also observed that the force testing error is systematically (but slightly) larger than the training error, which implies a slight over-fitting to the rather small dataset.

One can visualize this file by a simple Python script:

import numpy as np
import matplotlib.pyplot as plt

data = np.genfromtxt("lcurve.out", names=True)
for name in data.dtype.names[1:-1]:
    plt.plot(data['step'], data[name], label=name)
plt.legend()
plt.xlabel('Step')
plt.ylabel('Loss')
plt.xscale('symlog')
plt.yscale('log')
plt.grid()
plt.show()

During training, the model is saved in the TensorFlow model.ckpt* file every 10,000 steps, and the name of the last saved model is recorded in the checkpoint file.

When the training process is stopped abnormally, we can restart the training from the provided checkpoint by simply running

$ dp train  --restart model.ckpt  input.json

In the lcurve.out, you can see the training and testing errors, like

538000      3.12e-01       2.16e-01         6.84e-04         7.52e-04         1.38e-01         9.52e-02      4.1e-06
538000      3.12e-01       2.16e-01         6.84e-04         7.52e-04         1.38e-01         9.52e-02      4.1e-06
539000      3.37e-01       2.61e-01         7.08e-04         3.38e-04         1.49e-01         1.15e-01      4.1e-06
# step      rmse_val       rmse_trn       rmse_e_val       rmse_e_trn       rmse_f_val       rmse_f_trn           lr
530000      2.89e-01       2.15e-01         6.36e-04         5.18e-04         1.25e-01         9.31e-02      4.4e-06
531000      3.46e-01       3.26e-01         4.62e-04         6.73e-04         1.49e-01         1.41e-01      4.4e-06

Note that the input.json needs to be consistent with the previous one.

Freeze the model At the end of the training, the model parameters saved in TensorFlow’s checkpoint file should be frozen as a model file that is usually ended with extension .pb.

Simply execute

$ dp freeze -o graph.pb

where -o means output, which is used to specify the name of the output model. On the screen, you can see

DEEPMD INFO    Restoring parameters from ./model.ckpt-1000000
DEEPMD INFO    1264 ops in the final graph

and it will output a model file named graph.pb in the current directory.

Compress the model The compressed DP model typically speeds up DP-based calculations by an order of magnitude faster, and consumes an order of magnitude less memory. For a detailed description please refer to the literature.

The graph.pb can be compressed in the following way:

$ dp compress -i graph.pb -o graph-compress.pb

where -i means input, which is used to import unfrozen model. On the screen you can see

DEEPMD INFO    stage 1: compress the model
DEEPMD INFO    built lr
DEEPMD INFO    built network
DEEPMD INFO    built training
DEEPMD INFO    initialize model from scratch
DEEPMD INFO    finished compressing
DEEPMD INFO    
DEEPMD INFO    stage 2: freeze the model
DEEPMD INFO    Restoring parameters from model-compression/model.ckpt
DEEPMD INFO    840 ops in the final graph

and it will output a model file named graph-compress.pb.

Test the model Users can check the quality of the trained model by running

$ dp test -m graph-compress.pb -s ../00.data/validation_data -n 40 -d results

where -m means model which is used to import the model file, -s means system which specifies the path of the test dataset, -n means number which specifies the number of frames to be tested, and -d means detail which writes the details of energy, force and virial to different files.

On the screen, users can see the information on the prediction errors of validation data

DEEPMD INFO    # number of test data    : 40 
DEEPMD INFO    Energy RMSE              : 3.168050e-03 eV
DEEPMD INFO    Energy RMSE/Natoms       : 6.336099e-04 eV
DEEPMD INFO    Force  RMSE              : 1.267645e-01 eV/A
DEEPMD INFO    Virial RMSE              : 2.494163e-01 eV
DEEPMD INFO    Virial RMSE/Natoms       : 4.988326e-02 eV
DEEPMD INFO    # ----------------------------------------------- 

and it will output files named results.e.out and results.f.out in the current directory.

Model application

Run MD with LAMMPS

Users can use the DP model for MD simulations. Now let’s switch to 02.lmp folder to check the necessary input files for running MD with LAMMPS.

$ cd ../02.lmp

Firstly, we soft-link the output model in the training directory to the current directory

$ ln -s ../01.train/graph-compress.pb
```sh
Then we have three files
```sh
$ ls
conf.lmp  graph-compress.pb  in.lammps

where conf.lmp gives the initial configuration of a gas phase methane MD simulation, and the file in.lammps is the LAMMPS input script. One may check in.lammps and finds that it is a rather standard LAMMPS input file for a MD simulation, with only two exception lines:

pair_style  graph-compress.pb
pair_coeff  * *

where the pair style deepmd is invoked and the model file graph-compress.pb is provided, which means the atomic interaction will be computed by the DP model that is stored in the file graph-compress.pb.

One may execute lammps in the standard way

$ lmp  -i  in.lammps

After waiting for a while, the MD simulation finishes, and the log.lammps and ch4.dump files are generated. They store thermodynamic information and the trajectory of the molecule, respectively.

Model inference

Users can use the python interface of DeePMD-kit for model inference.

For example, users can use the DP model to evaluate the energy of each frame in the LAMMPS trajectory. Now start an interactive python environment,

$ python

then execute the following commands:

import dpdata  
d=dpdata.System('ch4.dump', fmt = "lammps/dump", type_map = ["H", "C"])  
d1 = d.predict(dp = "./graph-compress.pb")  
print(d1["energies"])

and it will print the energy of each snapshot on the screen

Summary

Now, users have learned the basic usage of the DeePMD-kit. For further information, please refer to the recommended links.

GitHub website：https://github.com/deepmodeling/
Documentations：https://docs.deepmodeling.com/
Tutorials：https://tutorials.deepmodeling.com/
Papers：https://deepmodeling.com/blog/papers/deepmd-kit/

Using DP-GEN

This tutorial tell you how to use DP-GEN，for detail information, you can check DP-GEN Doc

Install DP-GEN

There are various methods to install DP-GEN. Users can choose a suitable method depending on the machine environment. The following is a detailed description of two easy methods.

Install with pip

Users can use pip to install the DP-GEN if their machine can be connected to the internet. One can install DP-GEN directly by

pip install dpgen

Install with source code

Users can use the source code to install the DP-GEN if their machine cannot be connected to the internet. One can download the source code of DP-GEN by

git clone https://github.com/deepmodeling/dpgen.git

then you can install DP-GEN easily by:

cd dpgen
pip install --user .

With this command, the dpgen executable is install to $HOME/.local/bin/dpgen. You may want to export the PATH by

export PATH=$HOME/.local/bin:$PATH

Verify the installation

If the installation is successful, DP-GEN (dpgen) will be available to execute. To test the installation, you may execute

dpgen -h

the terminal will show the help information like

usage: dpgen [-h]
             {init_surf,init_bulk,auto_gen_param,init_reaction,run,run/report,collect,simplify,autotest,db} 
	     ...

dpgen is a convenient script that uses DeepGenerator to prepare initial data, drive DeepMDkit and analyze results. This script works based on several sub-commands with their own options. To see the options for the sub-commands, type "dpgen sub-command -h".

Hands-on tutorial for DP-GEN (v0.10.6)

Workflow of the DP-GEN

DeeP Potential GENerator (DP-GEN) is a package that implements a concurrent learning scheme to generate reliable DP models. Typically, the DP-GEN workflow contains three processes: init, run, and autotest.

init: generate the initial training dataset by first-principle calculations.
run: the main process of DP-GEN, in which the training dataset is enriched and the quality of the DP models is improved automatically.
autotest: calculate a simple set of properties and/or perform tests for comparison with DFT and/or empirical interatomic potentials.

This tutorial aims to help you quickly get command of the run process, so only a brief introduction to the init and auto-test processes is offered.

Example: a gas phase methane molecule

The following introduces the basic usage of the DP-GEN, taking a gas-phase methane molecule as an example.

Init

The initial dataset is used to train multiple (default 4) initial DP models and it can be generated in a custom way or in the standard way provided by DP-GEN.

Custom way

Performing ab-initio molecular dynamics (AIMD) simulations directly is a common custom way of generating initial data. The following suggestions are given to users who generate initial data via AIMD simulation:

Performing AIMD simulations at higher temperatures.
Start AIMD simulations from several (as many as possible) unrelated initial configurations.
Save snapshots from AIMD trajectories at a time interval to avoid sampling highly-related configurations.

Standard way of DP-GEN

For block materials, the initial data can be generated using DP-GE’s init_bulk method. In the init_bulk method, the given configuration is initially relaxed by ab-initio calculation and subsequently scaled or perturbed. Next, these scaled or perturbed configurations are used to start small-scale AIMD simulations, and the AIMD format data is finally converted to the data format required by DeePMD-kit. Basically, init_bulk can be divided into four parts:

Relax in folder 00.place_ele
Perturb and scale in folder 01.scale_pert
Run a short AIMD in folder 02.md
Collect data in folder 02.md.

For surface systems, the initial data can be generated using DP-GE’s init_surf method. Basically init_surf can be divided into two parts:

Build a specific surface in folder 00.place_ele
Perturb and scale in folder 01.scale_pert

Above steps are carried out automatically when generating the initial data in the standard way of DP-GEN. Users only need to prepare the input files for ab-initio calculation and DP-GEN (param.json and machine.json).

When generating the initial data for the block materials in the standard way, execute the following command:

$ dpgen init_bulk param.json machine.json

For surface systems, execute

$ dpgen init_surf param.json machine.json

A detailed description for preparing initial data in the standard way can be found at ‘Init’ Section of the DP-GEN’s documentation.

Initial data of this tutorial

In this tutorial, we take a gas-phase methane molecule as an example. We have prepared initial data in dpgen_example/init. Now download the dpgen_example and uncompress it:

wget https://dp-public.oss-cn-beijing.aliyuncs.com/community/dpgen_example.tar.xz
tar xvf dpgen_example.tar.xz

Go to and check the dpgen_example folder

$ cd dpgen_example  
$ ls  
init run

Folder init contains the initial data generated by AIMD simulations.
Folder run contains input files for the run process.

First, check the init folder with the tree command.

$ tree init -L 2

On the screen, you can see

init  
├── CH4.POSCAR  
├── CH4.POSCAR.01x01x01  
│ ├── 00.place_ele  
│ ├── 01.scale_pert
│ ├── 02.md  
│ └── param.json  
├── INCAR_methane.md  
├── INCAR_methane.rlx  
└── param.json

Folder CH4.POSCAR.01x01x01 contains the files generated by the DP-GEN init_bulk process.
INCAR_* and CH4.POSCAR are the standard INCAR and POSCAR files for VASP.
param.json is used to specify the details of the DP-GEN init_bulk process.

Note that POTCAR and machine.json are the same for init and run process, which can be found in the folder run.

Run

The run process contains a series of successive iterations, undertaken in order such as heating the system to certain temperature. Each iteration is composed of three steps: exploration, labeling, and training.

Input files

Firstly, we introduce the input files required for the run process of DP-GEN. We have prepared input files in dpgen_example/run

Now go into the dpgen_example/run.

$ cd dpgen_example/run
$ ls
INCAR_methane  machine.json  param.json  POTCAR_C  POTCAR_H

param.json is the settings for DP-GEN for the current task.
machine.json is a task dispatcher where the machine environment and resource requirements are set.
INCAR* and POTCAR* are the input file for the VASP package. All first-principle calculations share the same parameters as the one you set in param.json.

We can perform the run process as we expect by specifying the keywords in param.json and machine.json. A description of these keywords is given below.

param.json

The keywords in param.json can be split into 4 parts：

System and data: used to specify the atom types, initial data, etc.
Training: mainly used to specify tasks in the training step；
Exploration: mainly used to specify tasks in the labeling step；
Labeling: mainly used to specify tasks in the labeling step.

Here we introduce the main keywords in param.json, taking a gas-phase methane molecule as an example.

System and data

The system and data related keywords are given in the following:

"type_map": ["H","C"],
"mass_map": [1,12],
"init_data_prefix": "../",
"init_data_sys": ["init/CH4.POSCAR.01x01x01/02.md/sys-0004-0001/deepmd"],
"sys_configs_prefix": "../",
"sys_configs": [
     ["init/CH4.POSCAR.01x01x01/01.scale_pert/sys-0004-0001/scale-1.000/00000*/POSCAR"],
     ["init/CH4.POSCAR.01x01x01/01.scale_pert/sys-0004-0001/scale-1.000/00001*/POSCAR"]
],
"_comment": " that's all ",

Description of keywords:

Key	Type	Description
“type_map”	list	Atom types
“mass_map”	list	Standard atom weights.
“init_data_prefix”	str	Prefix of initial data directories
“init_data_sys”	list	Directories of initial data. You may use either the absolute or relative path here.
“sys_configs_prefix”	str	Prefix of sys_configs
“sys_configs”	list	Containing directories of structures to be explored in iterations. Wildcard characters are supported here.

Description of example:

The system related keys specify the basic information about the system. “type_map” gives the atom types, i.e. “H” and “C”. “mass_map” gives the standard atom weights, i.e. “1” and “12”.

The data related keys specify the init data for traning initial DP models and structures used for model_devi calculations. “init_data_prefix” and “init_data_sys” specify the location of the init data. “sys_configs_prefix” and “sys_configs” specify the location of the structures. Here, the init data is provided at “…… /init/CH4.POSCAR.01x01x01/02.md/sys-0004-0001/deepmd”. These structures are divided into two groups and provided at “……/init/CH4.POSCAR.01x01x01/01.scale_pert/sys-0004-0001/scale- 1.000/00000*/POSCAR” and “……/init/CH4.POSCAR.01x01x01/01.scale_pert/sys-0004-0001/scale- 1.000/00001*/POSCAR”.

Training

The training related keywords are given in the following:

"numb_models": 4,
"default_training_param": {
     "model": {
         "type_map": ["H","C"],
         "descriptor": {
             "type": "se_a",
             "sel": [16,4],
             "rcut_smth": 0.5,
             "rcut": 5.0,
             "neuron": [120,120,120],
             "resnet_dt": true,
             "axis_neuron": 12,
             "seed": 1
        },
         "fitting_net": {
             "neuron": [25,50,100],
             "resnet_dt": false,
             "seed": 1
         }
     },
     "learning_rate": {
         "type": "exp",
         "start_lr": 0.001,
         "decay_steps": 5000
     },
     "loss": {
         "start_pref_e": 0.02,
         "limit_pref_e": 2,
         "start_pref_f": 1000,
         "limit_pref_f": 1,
         "start_pref_v": 0.0,
         "limit_pref_v": 0.0
     },
     "training": {
         "stop_batch": 400000,
         "disp_file": "lcurve.out",
         "disp_freq": 1000,
         "numb_test": 4,
         "save_freq": 1000,
         "save_ckpt": "model.ckpt",
         "disp_training": true,
         "time_training": true,
         "profiling": false,
         "profiling_file": "timeline.json",
         "_comment": "that's all"
     }
 },

Description of keywords:

Key	Type	Description
“numb_models”	int	Number of models to be trained in 00.train.
“default_training_param”	dict	Training parameters for deepmd-kit.

Description of example:

The training related keys specify the details of training tasks. “numb_models” specifies the number of models to be trained. “default_training_param” specifies the training parameters for DeePMD-kit. Here, 4 DP models will be trained.

The training part of DP-GEN is performed by DeePMD-kit, so the keywords here are the same as those of DeePMD-kit and will not be explained here. A detailed explanation of those keywords can be found at DeePMD-kit’s documentation.

Exploration

The exploration related keywords are given in the following:

"model_devi_dt": 0.002,
"model_devi_skip": 0,
"model_devi_f_trust_lo": 0.05,
"model_devi_f_trust_hi": 0.15,
"model_devi_e_trust_lo": 10000000000.0,
"model_devi_e_trust_hi": 10000000000.0,
"model_devi_clean_traj": true,
"model_devi_jobs": [
     {"sys_idx": [0],"temps": [100],"press": [1.0],"trj_freq": 10,"nsteps": 300,"ensemble": "nvt","_idx": "00"},
     {"sys_idx": [1],"temps": [100],"press": [1.0],"trj_freq": 10,"nsteps": 3000,"ensemble": "nvt","_idx": "01"}
],

Description of keywords:

Key	Type	Description
“model_devi_dt”	float	Timestep for MD
“model_devi_skip”	int	Number of structures skipped for fp in each MD
“model_devi_f_trust_lo”	float	Lower bound of forces for the selection. If List, should be set for each index in sys_configs, respectively.
“model_devi_f_trust_hi”	int	Upper bound of forces for the selection. If List, should be set for each index in sys_configs, respectively.
“model_devi_v_trust_hi”	float or list	Lower bound of virial for the selection. If List, should be set for each index in sys_configs, respectively. Should be used with DeePMD-kit v2.x.
“model_devi_v_trust_hi”	float or list	Upper bound of virial for the selection. If List, should be set for each index in sys_configs, respectively. Should be used with DeePMD-kit v2.x.
“model_devi_clean_traj”	bool or int	If the type of model_devi_clean_traj is boolean type then it denotes whether to clean traj folders in MD since they are too large. If it is Int type, then the most recent n iterations of traj folders will be retained, others will be removed.
“model_devi_jobs”	list	Settings for exploration in 01.model_devi. Each dict in the list corresponds to one iteration. The index of model_devi_jobs exactly accord with the index of iterations
“sys_idx”	List of integer	Systems to be selected as the initial structure of MD and be explored. The index corresponds exactly to the “sys_configs”.
“temps”	list	Temperature (K) in MD
“press”	list	Pressure (Bar) in MD
“trj_freq”	int	Frequency of trajectory saved in MD.
“nsteps”	int	Running steps of MD.
“ensembles”	str	Determining which ensemble used in MD, options include “npt” and “nvt”.

Description of example:

The exploration related keys specify the details of exploration tasks. Here, MD simulations are performed at the temperature of 100 K and the pressure of 1.0 Bar with an integrator time of 2 fs under the nvt ensemble. Two iterations are set in “model_devi_jobs”. MD simulations are run for 300 and 3000 time steps with the first and second groups of structures in “sys_configs” in 00 and 01 iterations. We choose to save all structures generated in MD simulations and have set "trj_freq" as 10, so 30 and 300 structures are saved in 00 and 01 iterations. If the “max_devi_f” of saved structure falls between 0.05 and 0.15, DP-GEN will treat the structure as a candidate. We choose to clean traj folders in MD since they are too large. If you want to save the most recent n iterations of traj folders, you can set “model_devi_clean_traj” to be an integer.

Labeling

The labeling related keywords are given in the following:

"fp_style": "vasp",
"shuffle_poscar": false,
"fp_task_max": 20,
"fp_task_min": 5,
"fp_pp_path": "./",
"fp_pp_files": ["POTCAR_H","POTCAR_C"],
"fp_incar": "./INCAR_methane"

Description of keywords:

Key	Type	Description
“fp_style”	String	Software for First Principles. Options include “vasp”, “pwscf”, “siesta” and “gaussian” up to now.
“shuffle_poscar”	Boolean
“fp_task_max”	Integer	Maximum of structures to be calculated in 02.fp of each iteration.
“fp_task_min”	Integer	Minimum of structures to calculate in 02.fp of each iteration.
“fp_pp_path”	String	Directory of psuedo-potential file to be used for 02.fp exists.
“fp_pp_files”	List of string	Psuedo-potential file to be used for 02.fp. Note that the order of elements should correspond to the order in type_map.
“fp_incar”	String	Input file for VASP. INCAR must specify KSPACING and KGAMMA.

Description of example:

The labeling related keys specify the details of labeling tasks. Here, a minimum of 1 and a maximum of 20 structures will be labeled using the VASP code with the INCAR provided at “……/INCAR_methane” and POTCAR provided at “……/methane/POTCAR” in each iteration. Note that the order of elements in POSCAR and POTCAR should correspond to the order in type_map.

machine.json

Each iteration in the run process of DP-GEN is composed of three steps: exploration, labeling, and training. Accordingly, machine.json is composed of three parts: train, model_devi, and fp. Each part is a list of dicts. Each dict can be considered as an independent environment for calculation.

In this section, we will show you how to perform the training step at a local workstation, model_devi step at a local Slurm cluster, and fp step at a remote PBS cluster using the new DPDispatcher (the value of keyword “api_version” is larger than or equal to 1.0). For each step, three types of keys are needed:

Command: provides the command used to execute each step.
Machine: specifies the machine environment (local workstation, local or remote cluster, or cloud server).
Resources: specify the number of groups, nodes, CPU, and GPU; enable the virtual environment.

Performing the training step at a local workstation

In this example, we perform the training step on a local workstation.

"train": [
    {
      "command": "dp",
      "machine": {
        "batch_type": "Shell",
        "context_type": "local",
        "local_root": "./",
        "remote_root": "/home/user1234/work_path"
      },
      "resources": {
        "number_node": 1,
        "cpu_per_node": 4,
        "gpu_per_node": 1,
        "group_size": 1,
        "source_list": ["/home/user1234/deepmd.env"]
      }
    }
  ],

Description of keywords:

Key	Type	Description
“command”	String	A command to be executed of this task.
“machine”	dict	The definition of machine.
“batch_type”	str	The batch job system type.
“context_type”	str	The connection used to remote machine.
“local_root”	str	The dir where the tasks and relating files locate.
“remote_root”	str	The dir where the tasks are executed on the remote machine.
“machine”	dict	The definition of resources.
“number_node”	int	The number of node need for each job.
“cpu_per_node”	int	cpu numbers of each node assigned to each job.
“gpu_per_node”	int	gpu numbers of each node assigned to each job.
“group_size”	int	The number of tasks in a job.
“source_list”	str	The dir where the tasks are executed on the remote machine.

Description of example:

The “command” for the training tasks in the DeePMD-kit is “dp”.

In machine parameters, “batch_type” specifies the type of job scheduling system. If there is no job scheduling system, we can use the “Shell” to perform the task. “context_type” specifies the method of data transfer, and “local” means copying and moving data via local file storage systems (e.g. cp, mv, etc.). In DP-GEN, the paths of all tasks are automatically located and set by the software, and therefore “local_root” is always set to “. /”. The input file for each task will be sent to the “remote_root” and the task will be performed there, so we need to make sure that the path exists.

In the resources parameter, “number_node”, “cpu_per_node”, and “gpu_per_node” specify the number of nodes, the number of CPUs, and the number of GPUs required for a task respectively. “group_size”, which needs to be highlighted, specifies how many tasks will be packed into a group. In the training tasks, we need to train 4 models. If we only have one GPU, we can set the “group_size” to 4. If “group_size” is set to 1, 4 models will be trained on one GPU at the same time, as there is no job scheduling system. Finally, the environment variables can be activated by “source_list”. In this example, “source /home/user1234/deepmd.env” is executed before “dp” to load the environment variables necessary to perform the training task.

Perform the model_devi step at a local Slurm cluster

In this example, we perform the model_devi step at a local Slurm workstation.

"model_devi": [
    {
      "command": "lmp",
      "machine": {
       "context_type": "local",
        "batch_type": "Slurm",
        "local_root": "./",
        "remote_root": "/home/user1234/work_path"
      },
      "resources": {
        "number_node": 1,
        "cpu_per_node": 4,
        "gpu_per_node": 1,
        "queue_name": "QueueGPU",
        "custom_flags" : ["#SBATCH --mem=32G"],
        "group_size": 10,
        "source_list": ["/home/user1234/lammps.env"]
      }
    }
],

Description of keywords:

Key	Type	Description
“queue_name”	String	The queue name of batch job scheduler system.
“custom_flags”	String	The extra lines pass to job submitting script header.

Description of example:

The “command” for the model_devi tasks in the LAMMPS is “lmp”.

In the machine parameter, we specify the type of job scheduling system by changing the “batch_type” to “Slurm”.

In the resources parameter, we specify the name of the queue to which the task is submitted by adding “queue_name”. We can add additional lines to the calculation script via the “custom_flags”. In the model_devi steps, there are frequently many short tasks, so we usually pack multiple tasks (e.g. 10) into a group for submission. Other parameters are similar to that of the local workstation.

Perform the fp step in a remote PBS cluster

In this example, we perform the fp step at a remote PBS cluster that can be accessed via SSH.

"fp": [
    {
      "command": "mpirun -n 32 vasp_std",
      "machine": {
       "context_type": "SSHContext",
        "batch_type": "PBS",
        "local_root": "./",
        "remote_root": "/home/user1234/work_path",
        "remote_profile": {
          "hostname": "39.xxx.xx.xx",
          "username": "user1234"
         }
      },
      "resources": {
        "number_node": 1,
        "cpu_per_node": 32,
        "gpu_per_node": 0,
        "queue_name": "QueueCPU",
        "group_size": 5,
        "source_list": ["/home/user1234/vasp.env"]
      }
    }
],

Description of keywords:

Key	Type	Description
“remote_profile”	dict	The information used to maintain the connection with remote machine.
“hostname”	str	hostname or ip of ssh connection.
“username”	str	username of target linux system.

Description of example:

VASP code is used for fp tasks and mpi is used for parallel computing, so “mpirun -n 32” is added to specify the number of parallel threads.

In the machine parameter, “context_type” is modified to “SSHContext” and “batch_type” is modified to “PBS”. It is worth noting that “remote_root” should be set to an accessible path on the remote PBS cluster. “remote_profile” is added to specify the information used to connect the remote cluster, including hostname, username, password, port, etc.

In the resources parameter, we set “gpu_per_node” to 0 since it is cost-effective to use the CPU for VASP calculations.

Start run process

Once param.json and machine.json have been prepared, we can run DP-GEN easily by:

$ dpgen run param.json machine.json

Results analysis

Users need to know the output files of the run process and the information they contain. After successfully executing the above command, we can find that a folder and two files are generated automatically in dpgen_example/run.

$ ls 
dpgen.log  INCAR_methane  iter.000000  machine.json  param.json  record.dpgen 

iter.000000 contains the main results that DP-GEN generates in the first iteration.
record.dpgen records the current stage of the run process.
dpgen.log includes time and iteration information. When the first iteration is completed, the folder structure of iter.000000 is like this:

$ tree iter.000000/ -L 1
./iter.000000/
├── 00.train
├── 01.model_devi
└── 02.fp

00.train: several (default 4) DP models are trained on existing data.
01.model_devi: new configurations are generated using the DP models obtained in 00.train.
02.fp: first-principles calculations are performed on the selected configurations and the results are converted into training data.

00.train First, we check the folder iter.000000/ 00.train.

$ tree iter.000000/00.train -L 1
./iter.000000/00.train/
├── 000
├── 001
├── 002
├── 003
├── data.init -> /root/dpgen_example
├── data.iters
├── graph.000.pb -> 000/frozen_model.pb
├── graph.001.pb -> 001/frozen_model.pb
├── graph.002.pb -> 002/frozen_model.pb
└── graph.003.pb -> 003/frozen_model.pb

Folder 00x contains the input and output files of the DeePMD-kit, in which a model is trained.
graph.00x.pb , linked to 00x/frozen.pb, is the model DeePMD-kit generates. The only difference between these models is the random seed for neural network initialization. We may randomly select one of them, like 000.

$ tree iter.000000/00.train/000 -L 1
./iter.000000/00.train/000
├── checkpoint
├── frozen_model.pb
├── input.json
├── lcurve.out
├── model.ckpt-400000.data-00000-of-00001
├── model.ckpt-400000.index
├── model.ckpt-400000.meta
├── model.ckpt.data-00000-of-00001
├── model.ckpt.index
├── model.ckpt.meta
└── train.log

input.json is the settings for DeePMD-kit for the current task.
checkpoint is used for restart training.
model.ckpt* are model related files.
frozen_model.pb is the frozen model.
lcurve.out records the training accuracy of energies and forces.
train.log includes version, data, hardware information, time, etc.

01.model_devi Then, we check the folder iter.000000/ 01.model_devi.

$ tree iter.000000/01.model_devi -L 1
./iter.000000/01.model_devi/
├── confs
├── graph.000.pb -> /root/dpgen_example/run/iter.000000/00.train/graph.000.pb
├── graph.001.pb -> /root/dpgen_example/run/iter.000000/00.train/graph.001.pb
├── graph.002.pb -> /root/dpgen_example/run/iter.000000/00.train/graph.002.pb
├── graph.003.pb -> /root/dpgen_example/run/iter.000000/00.train/graph.003.pb
├── task.000.000000
├── task.000.000001
├── task.000.000002
├── task.000.000003
├── task.000.000004
├── task.000.000005
├── task.000.000006
├── task.000.000007
├── task.000.000008
└── task.000.000009

Folder confs contains the initial configurations for LAMMPS MD converted from POSCAR you set in “sys_configs” of param.json.
Folder task.000.00000x contains the input and output files of the LAMMPS. We may randomly select one of them, like task.000.000001.

$ tree iter.000000/01.model_devi/task.000.000001
./iter.000000/01.model_devi/task.000.000001
├── conf.lmp -> ../confs/000.0001.lmp
├── input.lammps
├── log.lammps
├── model_devi.log
└── model_devi.out

conf.lmp, linked to 000.0001.lmp in folder confs, serves as the initial configuration of MD.
input.lammps is the input file for LAMMPS.
model_devi.out records the model deviation of concerned labels, energy and force, in MD. It serves as the criterion for selecting which structures and doing first-principle calculations.

By head model_devi.out, you will see:

$ head -n 5 ./iter.000000/01.model_devi/task.000.000001/model_devi.out
 #  step max_devi_v     min_devi_v     avg_devi_v     max_devi_f     min_devi_f     avg_devi_f 
 0     1.438427e-04   5.689551e-05   1.083383e-04   8.835352e-04   5.806717e-04   7.098761e-04
10     3.887636e-03   9.377374e-04   2.577191e-03   2.880724e-02   1.329747e-02   1.895448e-02
20     7.723417e-04   2.276932e-04   4.340100e-04   3.151907e-03   2.430687e-03   2.727186e-03
30     4.962806e-03   4.943687e-04   2.925484e-03   5.866077e-02   1.719157e-02   3.011857e-02

Now we’ll concentrate on max_devi_f. Recall that we’ve set "trj_freq" as 10, so every 10 steps the structures are saved. Whether to select the structure depends on its "max_devi_f". If it falls between "model_devi_f_trust_lo" (0.05) and "model_devi_f_trust_hi" (0.15), DP-GEN will treat the structure as a candidate. Here, only the 30th structure will be selected, whose "max_devi_f" is 5.866077e e-02.

02.fp Finally, we check the folder iter.000000/ 02.fp.

$ tree iter.000000/02.fp -L 1
./iter.000000/02.fp
├── data.000
├── task.000.000000
├── task.000.000001
├── task.000.000002
├── task.000.000003
├── task.000.000004
├── task.000.000005
├── task.000.000006
├── task.000.000007
├── task.000.000008
├── task.000.000009
├── task.000.000010
├── task.000.000011
├── candidate.shuffled.000.out
├── POTCAR.000
├── rest_accurate.shuffled.000.out
└── rest_failed.shuffled.000.out

POTCAR is the input file for VASP generated according to "fp_pp_files" of param.json.
candidate.shuffle.000.out records which structures will be selected from last step 01.model_devi. There are always far more candidates than the maximum you expect to calculate at one time. In this condition, DP-GEN will randomly choose up to "fp_task_max" structures and form the folder task.*.
rest_accurate.shuffle.000.out records the other structures where our model is accurate (“max_devi_f” is less than "model_devi_f_trust_lo", no need to calculate any more),
rest_failed.shuffled.000.out records the other structures where our model is too inaccurate (larger than "model_devi_f_trust_hi", there may be some error).
data.000: After first-principle calculations, DP-GEN will collect these data and change them into the format DeePMD-kit needs. In the next iteration’s 00.train, these data will be trained together as well as the initial data.

By cat candidate.shuffled.000.out | grep task.000.000001, you will see:

$ cat ./iter.000000/02.fp/candidate.shuffled.000.out | grep task.000.000001
iter.000000/01.model_devi/task.000.000001 190
iter.000000/01.model_devi/task.000.000001 130
iter.000000/01.model_devi/task.000.000001 120
iter.000000/01.model_devi/task.000.000001 150
iter.000000/01.model_devi/task.000.000001 280
iter.000000/01.model_devi/task.000.000001 110
iter.000000/01.model_devi/task.000.000001 30
iter.000000/01.model_devi/task.000.000001 230

The task.000.000001 30 is exactly what we have just found in 01.model_devi satisfying the criterion to be calculated again. After the first iteration, we check the contents of dpgen.log and record.dpgen.

$ cat dpgen.log
2022-03-07 22:12:45,447 - INFO : start running
2022-03-07 22:12:45,447 - INFO : =============================iter.000000==============================
2022-03-07 22:12:45,447 - INFO : -------------------------iter.000000 task 00--------------------------
2022-03-07 22:12:45,451 - INFO : -------------------------iter.000000 task 01--------------------------
2022-03-08 00:53:00,179 - INFO : -------------------------iter.000000 task 02--------------------------
2022-03-08 00:53:00,179 - INFO : -------------------------iter.000000 task 03--------------------------
2022-03-08 00:53:00,187 - INFO : -------------------------iter.000000 task 04--------------------------
2022-03-08 00:57:04,113 - INFO : -------------------------iter.000000 task 05--------------------------
2022-03-08 00:57:04,113 - INFO : -------------------------iter.000000 task 06--------------------------
2022-03-08 00:57:04,123 - INFO : system 000 candidate :     12 in    310   3.87 %
2022-03-08 00:57:04,125 - INFO : system 000 failed    :      0 in    310   0.00 %
2022-03-08 00:57:04,125 - INFO : system 000 accurate  :    298 in    310  96.13 %
2022-03-08 00:57:04,126 - INFO : system 000 accurate_ratio:   0.9613    thresholds: 1.0000 and 1.0000   eff. task min and max   -1   20   number of fp tasks:     12
2022-03-08 00:57:04,154 - INFO : -------------------------iter.000000 task 07--------------------------
2022-03-08 01:02:07,925 - INFO : -------------------------iter.000000 task 08--------------------------
2022-03-08 01:02:07,926 - INFO : failed tasks:      0 in     12    0.00 % 
2022-03-08 01:02:07,949 - INFO : failed frame:      0 in     12    0.00 % 

It can be found that 310 structures are generated in iter.000000, in which 12 structures are collected for first-principle calculations.

$ cat record.dpgen
0
1
2
3
4
5
6
7
8

Each line contains two numbers: the first is the index of iteration, and the second, ranging from 0 to 9, records which stage in each iteration is currently running.

Index of iterations	“Stage in each iteration “	Process
0	0	make_train
0	1	run_train
0	2	post_train
0	3	make_model_devi
0	4	run_model_devi
0	5	post_model_devi
0	6	make_fp
0	7	run_fp
0	8	post_fp

If the process of DP-GEN stops for some reason, DP-GEN will automatically recover the main process by record.dpgen. You may also change it manually for your purpose, such as removing the last iterations and recovering from one checkpoint. After all iterations, we check the structure of dpgen_example/run

$ tree ./ -L 2
./
├── dpgen.log
├── INCAR_methane
├── iter.000000
│   ├── 00.train
│   ├── 01.model_devi
│   └── 02.fp
├── iter.000001
│   ├── 00.train
│   ├── 01.model_devi
│   └── 02.fp
├── iter.000002
│   └── 00.train
├── machine.json
├── param.json
└── record.dpgen

and contents of dpgen.log.

$ cat cat dpgen.log | grep system
2022-03-08 00:57:04,123 - INFO : system 000 candidate :     12 in    310   3.87 %
2022-03-08 00:57:04,125 - INFO : system 000 failed    :      0 in    310   0.00 %
2022-03-08 00:57:04,125 - INFO : system 000 accurate  :    298 in    310  96.13 %
2022-03-08 00:57:04,126 - INFO : system 000 accurate_ratio:   0.9613    thresholds: 1.0000 and 1.0000   eff. task min and max   -1   20   number of fp tasks:     12
2022-03-08 03:47:00,718 - INFO : system 001 candidate :      0 in   3010   0.00 %
2022-03-08 03:47:00,718 - INFO : system 001 failed    :      0 in   3010   0.00 %
2022-03-08 03:47:00,719 - INFO : system 001 accurate  :   3010 in   3010 100.00 %
2022-03-08 03:47:00,722 - INFO : system 001 accurate_ratio:   1.0000    thresholds: 1.0000 and 1.0000   eff. task min and max   -1    0   number of fp tasks:      0

It can be found that 3010 structures are generated in iter.000001, in which no structure is collected for first-principle calculations. Therefore, the final models are not updated in iter.000002/00.train.

Simplify

When you have a dataset containing lots of repeated data, this step will help you simplify your dataset.Since dpgen simplify is proformed on a large dataset, only a simple demo will be provided in this part.

To learn more about simplify, you can refer to DPGEN’s Document Document of dpgen simplify parameters Document of dpgen simplify machine parameters

This demo can be download from dpgen/examples/simplify-MAPbI3-scan-lebesgue. You can find more example in dpgen.examples

In the example, data contains a simplistic data set based on MAPbI3-scan case. Since it has been greatly reduced, do not take it seriously. It is just a demo. simplify_example is the work path, which contains INCAR and templates for simplify.json and machine.json. You can use the command nohup dpgen simplify simplify.json machine.json 1>log 2>err & here to test if dpgen simplify can run normally.

Kindly reminder:

machine.json is supported by dpdispatcher 0.4.15, please check https://docs.deepmodeling.com/projects/dpdispatcher/en/latest/ to update the parameters according to your dpdispatcher version.
POTCAR should be prepared by the user.
Please check the path and files name and make sure they are correct.

Simplify can be used in Transfer Learning, see CaseStudies: Transfer-learning

Auto-test

The function, auto-test, is only for alloy materials to verify the accuracy of their DP model, users can calculate a simple set of properties and compare the results with those of a DFT or traditional empirical force field. DPGEN’s autotest module supports the calculation of a variety of properties, such as

00.equi:(default task) the equilibrium state；
01.eos: the equation of state；
02.elastic: the elasticity like Young’s module；
03.vacancy: the vacancy formation energy；
04.interstitial: the interstitial formation energy；
05.surf: the surface formation energy.

In this part, the Al-Mg-Cu DP potential is used to illustrate how to automatically test DP potential of alloy materials. Each auto-test task includes three stages:

make prepares all required calculation files and input scripts automatically;
run can help submit calculation tasks to remote calculation plantforms and when calculation tasks are completed, will collect results automatically;
post returns calculation results to local root automatically.

structure relaxation

step1-`make`

Prepare the following files in a separate folder.

├── machine.json
├── relaxation.json
├── confs
│   ├── mp-3034

IMPORTANT! The ID number, mp-3034, is in the line with Material Project ID for Al-Mg-Cu.

In order to harness the benefits of pymatgen combined with Material Project to generate files for calculation tasks by mp-ID automatically, you are supposed to add the API for Material Project in the .bashrc.

You can do that easily by running this command.

vim .bashrc
// add this line into this file, `export MAPI_KEY="your-api-key-for-material-projects"`

If you have no ideas about api-key for material projects, please refer to this link.

machine.json is the same with the one used in init and run. For more information about it, please check this link.
relaxtion.json

{
    "structures":         ["confs/mp-3034"],//in this folder, confs/mp-3034, required files and scripts will be generated automatically by `dpgen autotest make relaxation.json`
    "interaction": {
            "type":        "deepmd",
            "model":       "graph.pb",
            "in_lammps":   "lammps_input/in.lammps",
            "type_map":   {"Mg":0,"Al": 1,"Cu":2} //if you  calculate other materials, remember to modify element types here.
    },
    "relaxation": {
            "cal_setting":{"etol": 1e-12,
                           "ftol": 1e-6,
                           "maxiter": 5000,
                           "maximal": 500000,
                           "relax_shape":     true,
                           "relax_vol":       true}
    }
}

Run this command,

dpgen autotest make relaxation.json

and then corresponding files and scripts used for calculation will be generated automatically.

step2-`run`

nohup dpgen autotest run relaxation.json machine.json &

After running this command, structures will be relaxed.

step3-`post`

dpgen autotest post relaxation.json

property calculation

step1-`make`

The parameters used for property calculations are in property.json.

{
    "structures":       ["confs/mp-3034"],
    "interaction": {
        "type":          "deepmd",
        "model":         "graph.pb",
        "deepmd_version":"2.1.0",
        "type_map":     {"Mg":0,"Al": 1,"Cu":2}
    },
    "properties": [
        {
         "type":         "eos",
         "vol_start":    0.9,
         "vol_end":      1.1,
         "vol_step":     0.01
        },
        {
         "type":         "elastic",
         "norm_deform":  2e-2,
         "shear_deform": 5e-2
        },
        {
         "type":             "vacancy",
         "supercell":        [3, 3, 3],
         "start_confs_path": "confs"
        },
        {
         "type":         "interstitial",
         "supercell":   [3, 3, 3],
         "insert_ele":  ["Mg","Al","Cu"],
         "conf_filters":{"min_dist": 1.5},
         "cal_setting": {"input_prop": "lammps_input/lammps_high"}
        },
        {
         "type":           "surface",
         "min_slab_size":  10,
         "min_vacuum_size":11,
         "max_miller":     2,
         "cal_type":       "static"
        }
        ]
}

Run this command

dpgen autotest make property.json

step2-`run`

Run this command

nohup dpgen autotest run property.json machine.json &

step3-`post`

dpgen autotest post property.json

In the folder, you can use the command tree . -L 1 and then you can check results.

(base) ➜ mp-3034 tree . -L 1
.
├── dpdispatcher.log
├── dpgen.log
├── elastic_00
├── eos_00
├── eos_00.bk000
├── eos_00.bk001
├── eos_00.bk002
├── eos_00.bk003
├── eos_00.bk004
├── eos_00.bk005
├── graph_new.pb
├── interstitial_00
├── POSCAR
├── relaxation
├── surface_00
└── vacancy_00

01.eos: the equation of state；

(base) ➜ mp-3034 tree eos_00 -L 1
eos_00
├── 99c07439f6f14399e7785dc783ca5a9047e768a8_flag_if_job_task_fail
├── 99c07439f6f14399e7785dc783ca5a9047e768a8_job_tag_finished
├── 99c07439f6f14399e7785dc783ca5a9047e768a8.sub
├── backup
├── graph.pb -> ../../../graph.pb
├── result.json
├── result.out
├── run_1660558797.sh
├── task.000000
├── task.000001
├── task.000002
├── task.000003
├── task.000004
├── task.000005
├── task.000006
├── task.000007
├── task.000008
├── task.000009
├── task.000010
├── task.000011
├── task.000012
├── task.000013
├── task.000014
├── task.000015
├── task.000016
├── task.000017
├── task.000018
├── task.000019
└── tmp_log

The EOS calculation results are shown in eos_00/results.out file

(base) ➜ eos_00 cat result.out 
conf_dir: /root/1/confs/mp-3034/eos_00
 VpA(A^3)  EpA(eV)
075   -3.2727 
242   -3.2838 
410   -3.2935 
577   -3.3019 
745   -3.3090 
912   -3.3148 
080   -3.3195 
247   -3.3230 
415   -3.3254 
582   -3.3268 
750   -3.3273 
917   -3.3268 
085   -3.3256 
252   -3.3236 
420   -3.3208 
587   -3.3174 
755   -3.3134 
922   -3.3087 
090   -3.3034 
257   -3.2977 

02.elastic: the elasticity like Young’s module； The elastic calculation results are shown in elastic_00/results.out file

(base) ➜ elastic_00 cat result.out 
/root/1/confs/mp-3034/elastic_00
 124.32   55.52   60.56    0.00    0.00    1.09 
  55.40  125.82   75.02    0.00    0.00   -0.17 
  60.41   75.04  132.07    0.00    0.00    7.51 
   0.00    0.00    0.00   53.17    8.44    0.00 
   0.00    0.00    0.00    8.34   37.17    0.00 
   1.06   -1.35    7.51    0.00    0.00   34.43 
# Bulk   Modulus BV = 84.91 GPa
# Shear  Modulus GV = 37.69 GPa
# Youngs Modulus EV = 98.51 GPa
# Poission Ratio uV = 0.31

03.vacancy: the vacancy formation energy； The vacancy calculation results are shown in vacancy_00/results.out file

(base) ➜ vacancy_00 cat result.out 
/root/1/confs/mp-3034/vacancy_00
Structure:      Vac_E(eV)  E(eV) equi_E(eV)
[3, 3, 3]-task.000000: -10.489  -715.867 -705.378 
[3, 3, 3]-task.000001:   4.791  -713.896 -718.687 
[3, 3, 3]-task.000002:   4.623  -714.064 -718.687 

04.interstitial: the interstitial formation energy； The interstitial calculation results are shown in interstitial_00/results.out file

(base) ➜ vacancy_00 cat result.out 
/root/1/confs/mp-3034/vacancy_00
Structure:      Vac_E(eV)  E(eV) equi_E(eV)
[3, 3, 3]-task.000000: -10.489  -715.867 -705.378 
[3, 3, 3]-task.000001:   4.791  -713.896 -718.687 
[3, 3, 3]-task.000002:   4.623  -714.064 -718.687 

05.surf: the surface formation energy. The surface calculation results are shown in surface_00/results.out file

(base) ➜ surface_00 cat result.out  
/root/1/confs/mp-3034/surface_00
Miller_Indices:         Surf_E(J/m^2) EpA(eV) equi_EpA(eV)
[1, 1, 1]-task.000000:          1.230      -3.102   -3.327
[1, 1, 1]-task.000001:          1.148      -3.117   -3.327
[2, 2, 1]-task.000002:          1.160      -3.120   -3.327
[2, 2, 1]-task.000003:          1.118      -3.127   -3.327
[1, 1, 0]-task.000004:          1.066      -3.138   -3.327
[2, 1, 2]-task.000005:          1.223      -3.118   -3.327
[2, 1, 2]-task.000006:          1.146      -3.131   -3.327
[2, 1, 1]-task.000007:          1.204      -3.081   -3.327
[2, 1, 1]-task.000008:          1.152      -3.092   -3.327
[2, 1, 1]-task.000009:          1.144      -3.093   -3.327
[2, 1, 1]-task.000010:          1.147      -3.093   -3.327
[2, 1, 0]-task.000011:          1.114      -3.103   -3.327
[2, 1, 0]-task.000012:          1.165      -3.093   -3.327
[2, 1, 0]-task.000013:          1.137      -3.098   -3.327
[2, 1, 0]-task.000014:          1.129      -3.100   -3.327
[1, 0, 1]-task.000015:          1.262      -3.124   -3.327
[1, 0, 1]-task.000016:          1.135      -3.144   -3.327
[1, 0, 1]-task.000017:          1.113      -3.148   -3.327
[1, 0, 1]-task.000018:          1.119      -3.147   -3.327
[1, 0, 1]-task.000019:          1.193      -3.135   -3.327
[2, 0, 1]-task.000020:          1.201      -3.089   -3.327
[2, 0, 1]-task.000021:          1.189      -3.092   -3.327
[2, 0, 1]-task.000022:          1.175      -3.094   -3.327
[1, 0, 0]-task.000023:          1.180      -3.100   -3.327
[1, 0, 0]-task.000024:          1.139      -3.108   -3.327
[1, 0, 0]-task.000025:          1.278      -3.081   -3.327
[1, 0, 0]-task.000026:          1.195      -3.097   -3.327
[2, -1, 2]-task.000027:         1.201      -3.121   -3.327
[2, -1, 2]-task.000028:         1.121      -3.135   -3.327
[2, -1, 2]-task.000029:         1.048      -3.147   -3.327
[2, -1, 2]-task.000030:         1.220      -3.118   -3.327
[2, -1, 1]-task.000031:         1.047      -3.169   -3.327
[2, -1, 1]-task.000032:         1.308      -3.130   -3.327
[2, -1, 1]-task.000033:         1.042      -3.170   -3.327
[2, -1, 0]-task.000034:         1.212      -3.154   -3.327
[2, -1, 0]-task.000035:         1.137      -3.165   -3.327
[2, -1, 0]-task.000036:         0.943      -3.192   -3.327
[2, -1, 0]-task.000037:         1.278      -3.144   -3.327
[1, -1, 1]-task.000038:         1.180      -3.118   -3.327
[1, -1, 1]-task.000039:         1.252      -3.105   -3.327
[1, -1, 1]-task.000040:         1.111      -3.130   -3.327
[1, -1, 1]-task.000041:         1.032      -3.144   -3.327
[1, -1, 1]-task.000042:         1.177      -3.118   -3.327
[2, -2, 1]-task.000043:         1.130      -3.150   -3.327
[2, -2, 1]-task.000044:         1.221      -3.135   -3.327
[2, -2, 1]-task.000045:         1.001      -3.170   -3.327
[1, -1, 0]-task.000046:         0.911      -3.191   -3.327
[1, -1, 0]-task.000047:         1.062      -3.168   -3.327
[1, -1, 0]-task.000048:         1.435      -3.112   -3.327
[1, -1, 0]-task.000049:         1.233      -3.143   -3.327
[1, 1, 2]-task.000050:          1.296      -3.066   -3.327
[1, 1, 2]-task.000051:          1.146      -3.097   -3.327
[1, 0, 2]-task.000052:          1.192      -3.085   -3.327
[1, 0, 2]-task.000053:          1.363      -3.050   -3.327
[1, 0, 2]-task.000054:          0.962      -3.132   -3.327
[1, -1, 2]-task.000055:         1.288      -3.093   -3.327
[1, -1, 2]-task.000056:         1.238      -3.102   -3.327
[1, -1, 2]-task.000057:         1.129      -3.122   -3.327
[1, -1, 2]-task.000058:         1.170      -3.115   -3.327
[0, 0, 1]-task.000059:          1.205      -3.155   -3.327
[0, 0, 1]-task.000060:          1.188      -3.158   -3.327

Summary

Now, users have learned the basic usage of the DP-GEN. For further information, please refer to the recommended links.

GitHub website：https://github.com/deepmodeling/dpgen
Papers：https://deepmodeling.com/blog/papers/dpgen/

Practical-Guidelines-for-DP

Practical Guidelines for DP

Before starting a new Deep Potential (DP) project, we suggest people (especially those who are newbies) read the following context first to get some insights into what tools we can use, what kinds of risks and difficulties we may meet, and how we can advance a new DP project smoothly. The contexts are written focused on “local configurational space”, which is very useful for thinking, analyzing, and solving problems when handling a DP project. The contexts are divided into three main parts:

“Know the Toolbox Well” gives a brief introduction to the DP method and the correlated tools of DP-GEN (Deep Potential GENerator) and DP Library (Deep Potential Library) from the point of view of local configuration space.
“Know the Physical Nature of a System” discusses how to set the parameters according to the properties of a material, which may be helpful to newbies.
“Know the Boundaries of a Problem” tells how we can generate a DP model that meets our requirements in a most efficient way, or how we can cut our project into pieces and make the project easier to implement.

Know the Toolbox Well

Knowing the toolbox well means that you know what the tools are, what the tools can be used to do, and that you know the limitations of the tools and what kind of risks may happen when using these tools.

Deep Potential

Deep Potential (DP) is a method that fits interatomic potentials (potential energy surface, PES) by deep neural networks usually from datasets calculated by DFT-based methods. The related software is DeePMD-kit. The DP method is general, accurate, computationally efficient, and scalable, which is one of the most popular machine learning potential methods. Similar to other machine learning potentials, the central ideal of DP is that the total energy of a system can be divided into the summation of potential energies of constitute atoms, $E=∑_{i=0}^{n}E_i$, where the potential energy of each atom $E_i$ depends on its local atomic configurations. There is one deep neural network model for each atom to describe $E_i$. A DP model consists of two sets of neural networks. The first one is an embedding net, which is designed under symmetry-invariant constraints and encodes the local environment of an atom into descriptors. The second one is a fitting net, which maps the output of the embedding net into $E_i$. DP_Illustration

Configurational Space

Local configurational space includes spatial configurations and chemical configurations. For spatial configurations, we take carbon as an example. It can crystallize either into graphite or diamond. In graphite, each carbon atom is connected to the other three carbons by $\rm{sp}^2$ bonds, forming a planner structure. In diamond, each carbon is connected to the other four carbons by $\rm{sp}^3$ bonds. In addition, carbon can also form many other structures, e.g. amorphous structures, fullerenes. Except for these explicit differences in local atomic structures, small deviations of atoms from their equilibrium positions also belong to spatial configurations. For chemical configurations, we take an ideal BCC lattice as an example. In the simplest case, all the lattice sites are occupied by the same atom, e.g. Ti. In another case, the corners are occupied by Ti, and the inner centers are occupied by Al, which is an ordered TiAl compound in the B2 structure. Thus, local chemical environments around Ti changes. In a more complex case, all the sites may be randomly occupied by different atoms, e.g. high entropy alloy, which may result in a vast chemical configurational space. It should be emphasized that this partition is just conceptional for easy understanding. In reality, the spatial and chemical configurations are coupled together and cannot be explicitly divided.

Configuration

Sampling Methods

From the above part, we know that “all we need is to try our best to cover the local configurational space” when developing a new DP model. Here, sampling methods that are usually adopted in practice are introduced. Roughly, the methods can be divided into four categories: manual design, MD+MC sampling, structural search, and enhanced sampling.

Manual construction Though many fancy methods can be used to assist sampling, manual design is still one of the most important sampling methods, especially for defects. For example, the initial structures of interstitials, vacancies, stacking faults, dislocations, surfaces, grain boundaries almost always need to be constructed explicitly. In some other cases, the initial structures also need to be constructed by experts, e.g. interfaces between different materials, absorption structures, etc.
MD+MC simulations MD and MC simulations are effective methods in sampling local regions in the configuration space, which are the simplest ways to implement in practice. For example, the vibration of atoms around their equilibrium positions in solids (MD), local environment changing in liquids (MD), exchanging similar atoms in a solid solution (MD+MC). The space that can be covered by an MD/MC simulation depends on the temperature and time of the simulation. In practice, other accelerated MD methods may also be adopted to assist sampling more efficiently.
Structure search Structure search methods, e.g. CALYPSO, USPEX, are useful for exploring reasonable structures of those materials with strong directional bonds, (e.g. most ceramics, and some other inorganic non-metallic materials carbon, boron, phosphorus, etc.) or unknown structures at high pressures. In such cases, neither manual construction nor MD+MC sampling is effective. Keep in mind that we are trying to cover the configurational space. Therefore, during structure search, not only the metastable or stable structures are collected to enrich our dataset, but also those structures with not very high energies are needed.
Enhanced sampling Enhanced sampling is an effective method to sample rare events, which are usually adopted to sample saddle points around a PES. In a system, the probability of a configuration being sampled is $p \propto exp(-U/k_BT)$. Therefore, high energy states, e.g. phase transformations, transition states of reaction, can hardly be sampled by MD simulations. In enhanced sampling methods, bias potentials are usually added to flatten the PES, which then enhances sampling of high energy states.

Sampling_methods

Limitations and Risks

Thanks to the representation and fitting powers of machine learning models, machine learning potentials are more precise than conventional potentials. However, a coin always has two sides. There is a typical shortcoming, the low capability of extrapolation, for almost all machine learning potentials. As illustrated in the following figure, the model matches extremely well in regions covered by the dataset, while predicting wrong results in regions that have not been covered by the dataset. In this example, two atoms may be bound together un-physically due to the lack of repulsion force around the core region. Commonly, atom pairs that are extremely close to each other are not in the dataset during sampling. To avoid such unphysical binds, we can artificially design some repulsion potentials in this region, or add dimer structures into the dataset and let the model learn the repulsion by itself. Risks This is only a simple example to illustrate the risk of nonphysical phenomena that may happen in simulations if the coverage of the training dataset on the local configurational space is poor. By illustrating the risk here, we are not going to encourage people to cover the whole configurational space when training a DP model. Instead, in the section “Know the Boundaries of a Problem”, we encourage people to sample the region of the configurational space that the problem locates in. It means that training a DP model that is sufficiently accurate for the problem to be studied is OK. In principle, the configurational space may be too big to be sufficiently sampled in some cases. Then, “Training a universally robust DP model is not a trivial work if it is not impossible”.

DP-GEN

DP-GEN is the abbreviation of “Deep Potential GENerator”, which is an automatic DP model generator that is designed under the concurrent learning framework. In DP-GEN, the enrichment of the dataset and the improvement of the DP model are done concurrently. The software DP-GEN can automatically manage the whole process, including preparing job scripts (e.g. training models, exploring configurational space and examining the accuracy of the configurations, and labeling by DFT-based calculations), submitting jobs to clouds or other HPCs (high-performance clusters), and monitoring job statuses. The following figure illustrates a typical auto iteration process of DP-GEN. There are three typical processes in an iteration:

Training a group of DP models (usually four), which are then used to explore the configurational space and check whether the configurations can be well precited
Exploring configurational space by one of the DP models and examining the prediction accuracy of each configuration by comparing the prediction deviations between different DP models
Selecting some configurations with low prediction accuracy as candidates and labeling the candidate configurations by DFT-based calculations

Initially, a dataset with hundreds to thousands of DFT data is provided, and then auto iterations can be started and run continuously. The initial dataset can be generated by the DP-GEN software (e.g. by using “init_bulk”, “init_surf” modules, or even the “autotest” module), or generated by yourself. During the DP-GEN process, the coverage of the dataset on the configurational space is limited, especially in the first few iterations. As has been stated in the section of “Limitations and Risks”, the capability of a DP model trained on the dataset is also limited, which predicts a configuration accurately when the configuration is in the explored region (IR), not satisfactory when the configuration is slightly out of the boundaries of the explore region (BR), and wrong when the configuration is far away from the explored region (FR). In addition, when using a DP model to explore the configurational space (e.g. by MD), nonphysical configurations may come out when the configuration is in the FR region. To avoid selecting nonphysical configurations, only those configurations in the BR region are selected as candidates. Therefore, both lower and upper bounds of prediction errors are set to select candidate configurations. In practice, if there is a valid conventional interatomic potential, sampling can be done by using the conventional potential, which is much more robust and may get rid of nonphysical configurations. In this case, the upper bound of prediction error can be set to a relatively big one. The boundaries of the explored regions extend along with these iterations by changing sampling methods or parameters, e.g. increasing MD temperature and simulation time. DP-GEN

DP Library

DFT-based calculations are expansive and time-consuming. Therefore, we built the DP library to share the source DFT data to prevent waste due to recalculation, and we encourage people to contribute to it, enrich the datasets, and improve the DP models continuously. With this infrastructure, data covering different regions of the configurational space may be contributed by different researchers, as shown in the figure below. Fortunately, a DP model can be retrained and improved when the dataset is enriched. Finally, we may reach many DP models that are good enough for most of the concerned problems and can focus on applications. When we need a DP model, then we can follow the steps to check what should we do:

Check whether a trained model exists, and get the model directly from the DP library
If not, check whether any source DFT data from the DP library is valuable for you, and add some data and train a model by yourself
If neither a trained model nor any valuable data exist, start to generate the data and train the model from scratch
Contribute the source data and models to the DP library, if you are willing to

DP_LIB

Know the Physical Nature of a System

Many parameters need to be set when handling a DP project, e.g. distortion and displacement parameters for generating initial data, temperatures, pressures, simulation times, etc. when running MD simulations to explore the configurational space. Though we can copy some scripts from others and get the DP-GEN run without changing any parameters, it is not a good idea in practice. Knowing the physical nature of your system can help you to design the parameters, getting a better hands-on experience. The local shape of PES （potential energy surface） around a configurational space region depends on the related bond strengths. The following figure illustrates a spectrum of chemical bonds. The PES is gentle with a widespread in the configurational space for soft chemical bonds, while the PES is sharp with a localized shape for strong chemical bonds. bond_nature

Sharp regions: deep valleys in PES

the vibration of a single molecule
the vibration of atoms in a solid

Gentle regions: shallow pits in PES

movements of atoms or molecules in liquids
solid solutions

Barrier regions:

phase transformations
transition states of reactions

Take a molecular liquid as an example. The intra-molecular bonds will result in a sharp PES that is localized in the configurational space, where atoms vibrate around their equilibrium positions. The characteristic time of vibration is very small (~ fs). Therefore, the coverage of sampling on the configuration space may be sufficient in a short-time MD simulation. In contrast, the inter-molecular bonds will lead to a gentle PES that is spread widely in the configurational space, where molecules move around each other with a large characteristic time (~ ps). Adequate sampling in the configurational space needs long-time MD simulations, or many short-time MD simulations starting from different configurations. Similar ideas are also applicable to chemical configurational space. Take $\rm{Zr}_{1−x}{Hf}_xC$ as an example. It is well known that changing Zr-Hf will not change the energy significantly, which is corresponding to a gentle PES and need a lot of MC steps to sample the space. Instead, the energy of anti-site defects Zr-C or Hf-C is very high. Thus, it is possible that sampling anti-site defects is not necessary. PES For simplicity, we will take Al as an example to explain some ideas, which give the relationship between the physical properties of a material and DP-GEN parameters.

Some initial data should be generated first, for example, by using the “init_bulk” method provided in DP-GEN. In this method, we need to set the ranges of linear compression/expansion, lattice distortion, and atom displacement. Usually, for a typical solid, volume expansion from room temperature to its melting point is ~5%, which is approximately ~2% along each dimension. Therefore, setting the range of linear compression/expansion to ±2% can usually cover the boundaries of a solid well, except for the high-pressure region. Random lattice distortions can also be set to a similar value, e.g. [-3%, 3%] for each strain mode. Atom displacements may be set by referring to the bond length of the nearest neighbor (e.g. < 1%d with d being the bond length). Usually, setting to 0.01 Å is OK.
When running the auto iterations by DP-GEN, we usually sample the configurational space by MD simulations with increasing temperature and a set of pressures. The setting of temperatures can refer to the melting point of Al, Tm (~1000 K), while the setting of pressures can refer to the bulk modulus of Al, B (~80 GPa). For example, the temperatures may be set into four groups: [0.0Tm, 0.5Tm], [0.5Tm, 1.0Tm], [1.0Tm, 1.5Tm], and [1.5Tm, 2.0Tm]. In each group, a few temperatures may be selected, e.g. cutting [0.0Tm, 0.5Tm] into [0.0Tm, 0.1Tm, 0.2Tm, 0.3Tm, 0.4Tm] (In practice, 0.0Tm is useless). The pressure fluctuates during MD simulations, the magnitude of which may be ~1% of B in small systems. Then, setting the pressures being [0.00B, 0.03B, 0.06B, 0.09B] is usually sufficient. Adding -0.01B may be helpful for solids and does not set negative pressures for the liquid to avoid the risk of continuous expansion of the simulation box.
The bounds for selecting candidates (“trust_level_low” and “trust_level_high”) may depend on the strongest chemical bonds in the system, e.g. the values may be higher in molecular systems than in metals. In solids, the lower and upper bound are roughly 0.2 and 0.5 that of the RMSE (reduced mean square error) of forces, $\sqrt{\sum f_i^2} $. Therefore, the melting temperature or bulk modulus is also a good indication for the bounds, since the forces are proportional to these properties. For example, the “trust_level_low” of Al DP-GEN is 0.05 eV/Å, while the “trust_level_low” of W DP-GEN is 0.15 eV/Å, three times that of Al. The melting point of W (~ 3600 K) is also nearly three times that of Al (~ 1000 K). These values may be good indicators when you insert them into the spectrum shown above. However, these criteria may not suitable for molecular liquids, which need relatively higher values due to the strong intramolecular bonds, even though their melting points are low.

Know the Boundaries of a Problem

Keep in mind that “Training a universally robust DP model is not a trivial work if it is not impossible”. Usually, we only need one DP model that meets our requirements. For different tasks, the desired coverages of the dataset on configurational space are different. Following is an example to illustrate this point of view:

We are interested in the room temperature elastic properties of Al

We are further interested in the temperature dependence of elastic properties of Al

We are interested in the melting point of Al

We are further interested in the solidification process of Al

We are interested in defects (e.g. vacancies, interstitials, dislocations, surfaces, grain boundaries) of Al

Example_Al In the first case, only configurations that are with small distortions around the equilibrium state are needed when calculating elastic properties. Therefore, only data around the equilibrium solid-state of Al are necessary. Running a DP-GEN with iterations from 000 to 006 may be enough. Additionally, if we would like to know the temperature dependence of elastic properties, iterations from 007 to 024, which further sample high energy state of Al bulk (e.g. expanded state due to thermal expansion), should be added.
In the second case, all the iterations in the figure should be done, which samples both the solid-state and liquid state of Al. However, if we are concerned about the nucleation details of Al from the liquid. The dataset may be enough (or not enough) to describe the solid-liquid interface accurately. Usually, the dataset is enough, if bonding in the material is not highly directional, e.g. for most metals. Sometimes, it is not, when bonding in the material is highly directional. For example, Ga, Si, etc. Then, an enhanced sampling method should be coupled into the DP-GEN process to sample the rare events, e.g. nucleation. For example, enforce the simulation running along the dotted line in the figure back and forth to gather samples around the saddle point.
In the third case, additional DP-GEN processes based on defect configurations (e.g. vacancies, interstitials, dislocations, surfaces, grain boundaries) are necessary. Fortunately, some local atomic configurations around defects may be similar to some distorted lattice structures or amorphous structures. Therefore, it is not necessary to explicitly include all the defect configurations during sampling. It can be seen from this simple example that a majority of efforts can be saved if the boundary of a problem can be well defined. For example, if we are only concerned about elastic properties of Al. It is not necessary to sample melts or defects, even though melts and defects are always sampled when developing DP models for metals. In contrast, when developing DP models for compounds, especially for those compounds with complex structures, melts and defects are only sampled when necessary. Therefore, before getting started with a new problem, pay some time to think about where the boundaries of the problem are and how much configurational space should be covered. When we get a new project, instead of being too excited to wait for getting into practice, plaining some milestones that are easier to achieve may facilitate the implementation of the project. For example, if our final goal is to investigate defects of Al, we can also cut the whole problem into pieces, and reach the final goal in stages, e.g. as stages 1, 2, and 3 stated above. After each stage, we can get a milestone, and then proceed to the next stage smoothly.

Convergence-Test

Convergence Test

Usually, a DP model is fitted to DFT data. The quality of the DFT dataset determines the accuracy limitation of the DP model. Ideally, we would like to generate DFT data as accurately as we can, e.g. using infinitely large ENCUT and infinitesimal KSPACING during DFT calculations. However, it is impossible in practice. We need to choose finite values for ENCUT and KSPACING. In addition, we need to achieve an acceptable trade-off between accuracy and efficiency. The ENCUT cannot be too large and the KSPACING cannot be too small. When training a DP model, tens of thousands of DFT single-point calculations would be carried out, which is very expansive meanwhile time-consuming. For example, assuming 10000 DFT calculations in total, 3 hours per calculation using nodes with 20 cores per-node, 600000 core*hours in total are needed. This is the reason why we encourage users to generate DFT data collaboratively. Therefore, to ensure the data quality, the reliability of the final model, as well as the feasibility of the project, a convergence test should be done first. Three goals are expected to achieve through the test:

to choose appropriate values of parameters in DFT calculations, such as ENCUT and KSPACING, ensuring that the DFT calculations are converged
to acquire an evaluation of the temporal and financial consumption of the project, which factor matters in designing the data generation scheme
to obtain a prior perception of error ranges as well as the intrinsic value of DFT energy, force, and stress (Virial), which are useful in setting pre-factors (and possibly fine-tune schemes) when training a DP model, as well as the configuration selection criteria in data generation scheme through DP-GEN

Special care should be taken that, 1) the convergence of three kinds of DFT labels (i.e. the energy, force, and Virial) are usually asynchronous, concerning the tightening up of calculation parameters 2) the convergence tendencies are configuration-dependent.

Before the test, convergence criteria for DFT results need to be clarified. Users are recommended to 1) make sure which labels are (more) important to the target properties or simulations in the project, 2) if attainable, map the expected accuracy of the targeted properties to the tolerances for specific labels (in specific configurations), 3) follow the general empirical convergence criteria, such as official recommendations from the developers of the adopted DFT code (could be reasonably overwritten by the specific requirement of the project).
Representive configurations are recommended to be used in convergence tests. A basic rule is to capture the maximum error fluctuation using configurations corresponding to extreme cases, i.e., to conduct a limit test. The cell_size/cell_shape/atom_perturbation… of the configuration should be customized according to the demands of the targeted properties.
- For a simple example, when elastic properties are focused, the extreme cases in finite-difference calculations that provide the largest label_Virials (thus might also present the largest errors in Virials) may demand uniaxial compressions around +/-2% in longitude (norm deformation) or shear around +/-2% in cell_vector_projection (shear deformation). Consequently, 2% or more remarkably deformed configuration(s) might be good candidates to be tested.
- Please note that the relation that extreme_deformation gives extreme_labels is case-specific, but configurations near equilibrium states would generally loosen the convergence condition where the intrinsic values of forces and Virials are close to zero. After making a successful compromise in calculation parameters, the third goal of the convergence test could also be considered.
for training scheme:
- In DP, the loss function is: $L(p_\epsilon,p_f,p_\xi)=p_\epsilon(\Delta\epsilon)^2+\frac{p_f}{3N}\sum_{i=1}^{N}\sum_{j=1}^{3}(\Delta F_{ij})^2+\frac{p_\xi}{9}\sum_{k=1}^{9}(\Delta \xi_k)^2$

The first, second, and third part are corresponding to energy error, force error, and stress error (Virial error), respectively.

The facts are that different labels should physically be self-consistent as expected by the DP model in constructing the PES, and all (three) labels are counted in the loss function. However, three DFT labels are in practice obtained by different numerical methods relying on sophisticated artificial implementations that make their error ranges not self-consistent. Consequently, the accuracy in the focused label (according to the demands of the project) of the DP model could be sacrificed as the price in fitting the other labels with larger noises.
To avoid such unwanted cases, users could try adjusting the conservative weights in the loss function according to their needs and based on the result of the convergence test
- Empirically, the error in Virials is the hardest to converge in most cases when using VASP and would influence the training accuracy and effectiveness in energy (training effectiveness, some cases need 10 times more training steps to achieve the same energy accuracy when using non-zero pref_V)
for data generation:
- In DP-GEN, the configuration selection is implemented by the model deviation in forces (trust_f), thus the results in values of forces could help in setting the trust_f. Further details about model deviation and rules in setting trust_f please see related documents. Traditionally, convergence test in DFT calculations only tests on the total energy of an equilibrium cell, e.g. to obtain ENCUT and KSPACING values that achieve an error level in total energy less than 1 meV/atom. In comparison to traditional ways, there are two major differences to doing the convergence test here.
When generating a DFT dataset to train a DP model, we need to check the convergence of energy, force, and stress. Usually, the convergence criteria are set to 0.001. For energy and stress, 0.001 means the error is 0.001 eV/atom, while it is 0.001 eV/Å for the force. To transform a stress value into an energy value, the stress is multiplied by the volume of the simulation cell.
A highly distorted supercell instead of an undistorted unit cell close to equilibrium is always used in the convergence test. For example, for a solid phase, firstly expand the volume of an equilibrium supercell by ~5%; then randomly deform the supercell, e.g. with the strain components chosen as random values with an upper limit of 3%; and finally, randomly shift each atom by a random value (e.g. < 0.02 Å) in each dimension. Choosing a randomly distorted supercell to do a convergence test is because almost all the DFT samples are distorted supercells. Using an equilibrium undistorted unit cell usually overestimates the convergence level. During the convergence test of ENCUT, the value of KSPACING is fixed at a low value, e.g. 0.10 $\rm{Å^{-1}}$. The variations of total energy (e), atomic force (f), and Virial (v) of a given configuration are calculated concerning the change of ENCUT. The following figure illustrates a typical example of how the total energy, atomic force, and Virial converge with ENCUT. In the figure, the values of ENCUT = 1200 eV are taken as references. ENCUT > 900 eV can meet our requirement when taking 0.001 as the criteria. The first one is the results of HfC, while the second one is the results of TaC.

During the convergence test of KSPACING, the value of ENCUT is fixed at a high value, e.g. 1000 eV. The variations of total energy (e), atomic force (f), and Virial (v) of a given configuration are calculated with respect to the change of KSPACING. The following figure illustrates a typical example of how the total energy, atomic force, and Virial converge with KSPACING. In the figure, the values of KSPACING = 0.08 $\rmÅ^{-1}$ are taken as references. Different from ENCUT, where good convergence tendency can be obtained, the convergence of force and Virial in KSPACING in this system is poor. It is possible that even the value of 0.08 $\rmÅ^{-1}$ cannot reach the 0.001 criteria for force and Virial from this figure. Nevertheless, the energy convergence is good enough. All the KSPACING values may meet the 0.001 criteria. The first one is the results of HfC, while the second one is the results of TaC. kspacing_HfC kspacing_TaC

Overall, we can see that the accuracy of energy is the best among energy, atomic force, and Virial. In practice, the energy values are the most trustable. The accuracy of these arguments will influence the prediction accuracy of materials properties. For example, the following table shows the influence of convergence level on the prediction of properties. Since the energy converges well, the predictions on lattice parameters and cohesive energy are good, which do not vary significantly with KSPACING. However, the predictions on elastic constants depend strongly on the value of KSPACING, especially for TaC. As can be seen from the figure above, errors of TaC in force and stress are much higher than those of HfC. Therefore, the prediction accuracy of elastic constants of TaC is low. In practice, the smallest KSPACING value affordable is around 0.10 $\rm{Å^{-1}}$, since we need to do tens of thousands of DFT calculations. Therefore, knowing the convergence level of your DFT calculations is essential, especially when we cannot reach the 0.001 criteria.

Compound	KSPACING(1/Å)	a(Å)	E/atom(eV/atom)	C11(GPa)	C12(GPa)	C44(GPa)
HfC	0.15	4.6467	-10.5247	518.11	103.43	170.21
HfC	0.10	4.6467	-10.5247	519.44	102.12	175.52
HfC	0.05	4.6470	-10.5253	512.87	104.64	172.28
TaC	0.15	4.4782	-11.1065	590.20	193.68	171.34
TaC	0.10	4.4782	-11.1065	671.24	154.41	170.82
TaC	0.05	4.4785	-11.1039	708.46	134.13	175.99

Gas-Phase

Simulation of the oxidation of methane

Jinzhe Zeng, Liqun Cao, and Tong Zhu

This tutorial was adapted from: Jinzhe Zeng, Liqun Cao, Tong Zhu (2022), Neural network potentials, Pavlo O. Dral (Eds.), Quantum Chemistry in the Age of Machine Learning, Elsevier. Please cite the above chapter if you follow the tutorial.

In this tutorial, we will take the simulation of methane combustion as an example and introduce the procedure of DP-based MD simulation. All files needed in this section can be downloaded from tongzhugroup/Chapter13-tutorial. Besides DeePMD-kit (with LAMMPS), ReacNetGenerator should also be installed.

Step 1: Preparing the reference dataset

In the reference dataset preparation process, one also has to consider the expect accuracy of the final model, or at what QM level one should label the data. In this paper, the Gaussian software was used to calculate the potential energy and atomic forces of the reference data at the MN15/6-31G** level. The MN15 functional was employed because it has good accuracy for both multi-reference and single-reference systems, which is essential for our system as we have to deal with a lot of radicals and their reactions. Here we assume that the dataset is prepared in advance, which can be downloaded from tongzhugroup/Chapter13-tutorial.

Step 2. Training the Deep Potential (DP)

Before the training process, we need to prepare an input file called methane_param.json which contains the control parameters. The training can be done by the following command:

$ $deepmd_root/bin/dp train methane_param.json

There are several parameters we need to define in the methane_param.json file. The type_map refers to the type of elements included in the training, and the option of rcut is the cut-off radius which controls the description of the environment around the center atom. The type of descriptor is se_a in this example, which represents the DeepPot-SE model. The descriptor will decay smoothly from rcut_smth (R_on) to the cut-off radius rcut (R_off). Here rcut_smth and rcut are set to 1.0 Å and 6.0 Å respectively. The sel defines the maximum possible number of neighbors for the corresponding element within the cut-off radius. The options neuron in descriptor and fitting_net is used to determine the shape of the embedding neural network and the fitting network, which are set to (25, 50, 100) and (240, 240, 240) respectively. The value of axis_neuron represents the size of the embedding matrix, which was set to 12.

Step 3: Freeze the model

This step is to extract the trained neural network model. To freeze the model, the following command will be executed:

$ $deepmd_root/bin/dp freeze -o graph.pb

A file called graph.pb can be found in the training folder. Then the frozen model can be compressed:

$ $deepmd_root/bin/dp compress -i graph.pb -o graph_compressed.pb -t methane_param.json

Step 4: Running MD simulation based on the DP

The frozen model can be used to run reactive MD simulations to explore the detailed reaction mechanism of methane combustion. The MD engine is provided by the LAMMPS software. Here we use the same system from our previous work, which contains 100 methane and 200 oxygen molecules. The MD will be performed under the NVT ensemble at 3000 K for 1 ns. The LAMMPS program can be invoked by the following command:

$ $deepmd_root/bin/lmp -i input.lammps

The input.lammps is the input file that controls the MD simulation in detail, technique details can be found in the manual of LAMMPS. To use the DP, the pair_style option in this input should be specified as follows:

pair_style deepmd graph_compressed.pb 
pair_coeff * * 

Step 5: Analysis of the trajectory

After the simulation is done, we can use the ReacNetGenerator software which was developed in our previous study to extract the reaction network from the trajectory. All species and reactions in the trajectory will be put on an interactive web page where we can analyze them by mouse clicks. Eventually we should be able to obtain reaction networks that consistent with the following figure.

$ reacnetgenerator -i methane.lammpstrj -a C H O --dump

The initial stage of combustion

Fig: The initial stage of combustion. The figure is taken from this paper and more results can be found there.

Acknowledge

This work was supported by the National Natural Science Foundation of China (Grants No. 22173032, 21933010). J.Z. was supported in part by the National Institutes of Health (GM107485) under the direction of Darrin M. York. We also thank the ECNU Multifunctional Platform for Innovation (No. 001) and the Extreme Science and Engineering Discovery Environment (XSEDE), which is supported by National Science Foundation Grant ACI-1548562.56 (specifically, the resources EXPANSE at SDSC through allocation TG-CHE190067), for providing supercomputer time.

References

Jinzhe Zeng, Liqun Cao, Tong Zhu (2022), Neural network potentials, Pavlo O. Dral (Eds.), Quantum Chemistry in the Age of Machine Learning, Elsevier.
Jinzhe Zeng, Liqun Cao, Mingyuan Xu, Tong Zhu, John Z. H. Zhang, Complex reaction processes in combustion unraveled by neural network-based molecular dynamics simulation, Nature Communications, 2020, 11, 5713.
Frisch, M.; Trucks, G.; Schlegel, H.; Scuseria, G.; Robb, M.; Cheeseman, J.; Scalmani, G.; Barone, V.; Petersson, G.; Nakatsuji, H., Gaussian 16, revision A. 03. Gaussian Inc., Wallingford CT 2016.
Han Wang, Linfeng Zhang, Jiequn Han, Weinan E, DeePMD-kit: A deep learning package for many-body potential energy representation and molecular dynamics, Computer Physics Communications, 2018, 228, 178-184.
Aidan P. Thompson, H. Metin Aktulga, Richard Berger, Dan S. Bolintineanu, W. Michael Brown, Paul S. Crozier, Pieter J. in ‘t Veld, Axel Kohlmeyer, Stan G. Moore, Trung Dac Nguyen, Ray Shan, Mark J. Stevens, Julien Tranchida, Christian Trott, Steven J. Plimpton, LAMMPS - a flexible simulation tool for particle-based materials modeling at the atomic, meso, and continuum scales, Computer Physics Communications, 2022, 271, 108171.
Denghui Lu, Wanrun Jiang, Yixiao Chen, Linfeng Zhang, Weile Jia, Han Wang, Mohan Chen, DP Train, then DP Compress: Model Compression in Deep Potential Molecular Dynamics, 2021.
Jinzhe Zeng, Liqun Cao, Chih-Hao Chin, Haisheng Ren, John Z. H. Zhang, Tong Zhu, ReacNetGenerator: an automatic reaction network generator for reactive molecular dynamics simulations, Phys. Chem. Chem. Phys., 2020, 22 (2), 683–691.

Mg-Y_alloy

Mg-Y alloy system

Introduction

We will briefly analyze the candidate configurational space of a metallic system by taking Mg-based Mg-Y binary alloy as an example. In atomistic simulations on metallic materials, defects (point defects, dislocations, grain boundaries, etc.) are usually concerned. In order to meet the requirements in these simulations, the potential should have correct descriptions on:

Non-defective Mg and Mg-Y solid solutions with HCP, BCC, and FCC structures, and Mg-Y compounds (Mg24Y5, Mg2Y, MgY). For example, lattice parameters, cohesive energy, elastic properties. Since we are concerned with Mg-based alloys, the concentrations of Y in solid solutions are limited to be lower than 20 at. % (the highest solid solution concentration of Y in Mg is 3.75 at. %, as shown in the Mg-Y phase diagram). However, the local concentration of Y may be higher due to segregation, e.g. around dislocations or grain boundaries. 20 at. % of Y is close to the first stable Mg24Y5 compound in the Mg-rich end in Mg-Y phase diagram. NO NEED to explore the full concentration. This will save lots of time and money.
Typical defects in Mg-Y system, including, vancancies, intersititials, dislocations, twins, stacking faults, grain boundaries, and surfaces. For example, the formation energies of defects, the interaction between solid solutes and defects.
Armophous structures. Even though we are not concerned with Mg-Y melts, local high-energy regions in simulations may exhibit amorphous-like structures, e.g. high-energy grain boundaries at high temperatures.

In our work, we are concerned with the interaction between the solute atom Y and dislocations, grain boundaries, and evaluating the solute segregation effects on properties of these defects. Therefore, our final DP model should meet all the requirements listed above. In addition, the DP should also be stable to achieve long-time simulation at large scales, since the simulation time is as long as tens of nanoseconds in systems with tens of thousands of atoms. Thus, the DP should be stable enough to do the simulation.

Therefore, we divided the task into the following steps during the DP-GEN process:

Only perfect crystals of Mg, Y, Mg-Y solid solutes (HCP, BCC, and FCC) and Mg-Y compounds (Mg24Y5, Mg2Y, MgY) were considered. During this stage, hybrid MC/MD simulations with npt-iso, npt-aniso or npt-tri ensemble will be used to sample both changes in local chemical configurational space and in local spacial configurational space (vibrations of atoms from their equilibrium positions).
Defective structures, including interstitials, vacancies, stacking faults, and surfaces, were constructed and sampled with similar strategies as those in the first stage. The defective structures were only constructed for pure Mg and Mg-Y solid solutions. Usually, a dislocation model is too large to be calculated by DFT with high precision. Therefore, dislocations are not explicitly included. During this stage, the temperatures were limited to a value that was not too high, e.g. 0.5Tm (Tm is the melting point of Mg). For surface models, nvt ensembles were preferred. Therefore, artificial deformations can be applied to the simulation cells .
Amorphous (liquid) structures with different densities and Y concentrations were considered, which are obtained by burning the above structures to high temperatures at different pressures (to change their density). During this stage, hybrid MC/MD simulations with npt-iso ensemble will be used to ensure that the shape of a simulation cell is reasonable, since the cell shape of an amorphous may change to an unreasonable one in npt-aniso or npt-tri ensemble. Alternatively, hybrid MC/MD simulations with nvt ensemble can also be adopted, if amorphous structures can be well constructed.

Folder Structure

A typical folder structure of DP-GEN process can be set as following:

Mg-Y (The name of your project)

|—-pseudopotentials (pseudopotentials for elements, e.g. Mg: PAW_PBE, Y: PAW_PBE_sv)

|—-input_fp (input files of the DFT calculations, including single-point calculation, relaxation and MD simulation)

|—-init (generating initial data by DP-GEN “init_bulk” or “init_surf”, other initial DFT data generated by yourself or downloaded from the DP library )

|—-run (auto iteration folder for DP-GEN run)

|—-train_test (training and testing DP models after finished DP-GEN)

|—-autotest (test properties by DP-GEN autotest)

Preparing Files

By checking the DP library, we can find that there are some data for Mg substance. Therefore, these data are downloaded and taken as part of the inital data.

The examplory files are attached: init_bulk param.json, run param.json. machine-alo.json

{
   "stages" : [1,2,3,4],
   "elements":     ["Mg","Y"],
   "cell_type":    "bcc",
   "latt":     5.160,
   "super_cell":   [2, 2, 2],
   "from_poscar":	true,
   "from_poscar_path":  "./POSCAR",
   "potcars":      ["/data/wangyinan/wangyinan/Mg-Y/files/POTCAR_Mg","/data/wangyinan/wangyinan/Mg-Y/files/POTCAR_Y"],
   "relax_incar": "/data/wangyinan/wangyinan/Mg-Y/files/INCAR_rlx",
   "md_incar" : "/data/wangyinan/wangyinan/Mg-Y/files/INCAR_md",
   "skip_relax":   true,
   "scale":        [0.97, 1.00, 1.03],
   "pert_numb":    20,
   "md_nstep" :    5,
   "pert_box":     0.03,
   "pert_atom":    0.01,
   "coll_ndata":   5000,
   "type_map" :    ["Mg","Y"],   
   "_comment":     "that's all"
}

{
    "type_map": ["Mg","Y"],
    "mass_map": [24.3,89],
    
    "init_data_prefix": "/data/wangyinan/wangyinan/Mg-Y/init/",
    
    "init_data_sys": [
     
        "./Mg/init/high_pressure/mg.bcc.02x02x02/02.md/sys-0016/deepmd",
        "./Mg/init/high_pressure/mg.fcc.02x02x02/02.md/sys-0032/deepmd",
        "./Mg/init/high_pressure/mg.hcp.02x02x02/02.md/sys-0016/deepmd",
  
        "./Mg/iter.000000/02.fp/data.000",
        "./Mg/iter.000000/02.fp/data.008",
        "./Mg/iter.000000/02.fp/data.016",
        "./Mg/iter.000008/02.fp/data.000",
        "./Mg/iter.000008/02.fp/data.008",
        "./Mg/iter.000009/02.fp/data.001",
        "./Mg/iter.000009/02.fp/data.009",
        "./Mg/iter.000009/02.fp/data.017",
        "./Mg/iter.000011/02.fp/data.003",
        "./Mg/iter.000011/02.fp/data.019",
        "./Mg/iter.000012/02.fp/data.004",
        "./Mg/iter.000013/02.fp/data.005",
        "./Mg/iter.000014/02.fp/data.022",
        "./Mg/iter.000015/02.fp/data.007",
        "./Mg/iter.000016/02.fp/data.000",
        "./Mg/iter.000016/02.fp/data.008",
        "./Mg/iter.000016/02.fp/data.016",
        "./Mg/iter.000017/02.fp/data.009",
        "./Mg/iter.000018/02.fp/data.002",
        "./Mg/iter.000018/02.fp/data.010",
        "./Mg/iter.000019/02.fp/data.019",
        "./Mg/iter.000020/02.fp/data.004",
        "./Mg/iter.000020/02.fp/data.012",
        "./Mg/iter.000020/02.fp/data.020",
        "./Mg/iter.000021/02.fp/data.013",
        "./Mg/iter.000021/02.fp/data.021",
        "./Mg/iter.000022/02.fp/data.022",
        "./Mg/iter.000023/02.fp/data.007",
        "./Mg/iter.000024/02.fp/data.000",
        "./Mg/iter.000024/02.fp/data.008",
        "./Mg/iter.000024/02.fp/data.016",
        "./Mg/iter.000025/02.fp/data.001",
        "./Mg/iter.000025/02.fp/data.009",
        "./Mg/iter.000026/02.fp/data.002",
        "./Mg/iter.000026/02.fp/data.018",
        "./Mg/iter.000032/02.fp/data.000",
        "./Mg/iter.000034/02.fp/data.002",
        "./Mg/iter.000034/02.fp/data.010",
        "./Mg/iter.000034/02.fp/data.018",
        "./Mg/iter.000035/02.fp/data.003",
        "./Mg/iter.000035/02.fp/data.019",
        "./Mg/iter.000036/02.fp/data.004",
        "./Mg/iter.000036/02.fp/data.012",
        "./Mg/iter.000036/02.fp/data.020",
        "./Mg/iter.000037/02.fp/data.005",
        "./Mg/iter.000037/02.fp/data.013",
        "./Mg/iter.000037/02.fp/data.021",
        "./Mg/iter.000038/02.fp/data.006",
        "./Mg/iter.000038/02.fp/data.014",
        "./Mg/iter.000039/02.fp/data.007",
        "./Mg/iter.000039/02.fp/data.015",
        "./Mg/iter.000046/02.fp/data.046",
        "./Mg/iter.000047/02.fp/data.031",
        "./Mg/iter.000047/02.fp/data.047",
        "./Mg/iter.000049/02.fp/data.025",
        "./Mg/iter.000050/02.fp/data.026",
        "./Mg/iter.000050/02.fp/data.030",
        "./Mg/iter.000050/02.fp/data.046",
        "./Mg/iter.000051/02.fp/data.027",
        "./Mg/iter.000051/02.fp/data.031",
        "./Mg/iter.000051/02.fp/data.035",
        "./Mg/iter.000051/02.fp/data.039",
        "./Mg/iter.000051/02.fp/data.043",
        "./Mg/iter.000051/02.fp/data.047",
        "./Mg/iter.000052/02.fp/data.024",
        "./Mg/iter.000052/02.fp/data.028",
        "./Mg/iter.000052/02.fp/data.044",
        "./Mg/iter.000053/02.fp/data.025",
        "./Mg/iter.000053/02.fp/data.029",
        "./Mg/iter.000053/02.fp/data.033",
        "./Mg/iter.000053/02.fp/data.045",
        "./Mg/iter.000054/02.fp/data.026",
        "./Mg/iter.000054/02.fp/data.030",
        "./Mg/iter.000054/02.fp/data.034",
        "./Mg/iter.000054/02.fp/data.038",
        "./Mg/iter.000054/02.fp/data.042",
        "./Mg/iter.000054/02.fp/data.046",
        "./Mg/iter.000055/02.fp/data.027",
        "./Mg/iter.000055/02.fp/data.031",
        "./Mg/iter.000055/02.fp/data.035",
        "./Mg/iter.000055/02.fp/data.043",
        "./Mg/iter.000055/02.fp/data.047",

	"./Y/9R/1-0/deepmd",
	"./Y/9R/1-1/deepmd",
	"./Y/9R/1-10/deepmd",
	"./Y/9R/1-11/deepmd",
	"./Y/9R/1-12/deepmd",
	"./Y/9R/1-13/deepmd",
	"./Y/9R/1-14/deepmd",
	"./Y/9R/1-15/deepmd",
	"./Y/9R/1-16/deepmd",
	"./Y/9R/1-17/deepmd",
	"./Y/9R/1-18/deepmd",
	"./Y/9R/1-19/deepmd",
	"./Y/9R/1-2/deepmd",
	"./Y/9R/1-3/deepmd",
	"./Y/9R/1-4/deepmd",
	"./Y/9R/1-5/deepmd",
	"./Y/9R/1-6/deepmd",
	"./Y/9R/1-7/deepmd",
	"./Y/9R/1-8/deepmd",
	"./Y/9R/1-9/deepmd",
	"./Y/9R/2-0/deepmd",
	"./Y/9R/2-1/deepmd",
	"./Y/9R/2-10/deepmd",
	"./Y/9R/2-11/deepmd",
	"./Y/9R/2-12/deepmd",
	"./Y/9R/2-13/deepmd",
	"./Y/9R/2-14/deepmd",
	"./Y/9R/2-15/deepmd",
	"./Y/9R/2-16/deepmd",
	"./Y/9R/2-17/deepmd",
	"./Y/9R/2-18/deepmd",
	"./Y/9R/2-19/deepmd",
	"./Y/9R/2-2/deepmd",
	"./Y/9R/2-3/deepmd",
	"./Y/9R/2-4/deepmd",
	"./Y/9R/2-5/deepmd",
	"./Y/9R/2-6/deepmd",
	"./Y/9R/2-7/deepmd",
	"./Y/9R/2-8/deepmd",
	"./Y/9R/2-9/deepmd",
	"./Y/9R/3-0/deepmd",
	"./Y/9R/3-1/deepmd",
	"./Y/9R/3-10/deepmd",
	"./Y/9R/3-11/deepmd",
	"./Y/9R/3-12/deepmd",
	"./Y/9R/3-13/deepmd",
	"./Y/9R/3-14/deepmd",
	"./Y/9R/3-15/deepmd",
	"./Y/9R/3-16/deepmd",
	"./Y/9R/3-17/deepmd",
	"./Y/9R/3-18/deepmd",
	"./Y/9R/3-19/deepmd",
	"./Y/9R/3-2/deepmd",
	"./Y/9R/3-3/deepmd",
	"./Y/9R/3-4/deepmd",
	"./Y/9R/3-5/deepmd",
	"./Y/9R/3-6/deepmd",
	"./Y/9R/3-7/deepmd",
	"./Y/9R/3-8/deepmd",
	"./Y/9R/3-9/deepmd",
	"./Y/dhcp/1-0/deepmd",
	"./Y/dhcp/1-1/deepmd",
	"./Y/dhcp/1-10/deepmd",
	"./Y/dhcp/1-11/deepmd",
	"./Y/dhcp/1-12/deepmd",
	"./Y/dhcp/1-13/deepmd",
	"./Y/dhcp/1-14/deepmd",
	"./Y/dhcp/1-15/deepmd",
	"./Y/dhcp/1-16/deepmd",
	"./Y/dhcp/1-17/deepmd",
	"./Y/dhcp/1-18/deepmd",
	"./Y/dhcp/1-19/deepmd",
	"./Y/dhcp/1-2/deepmd",
	"./Y/dhcp/1-3/deepmd",
	"./Y/dhcp/1-4/deepmd",
	"./Y/dhcp/1-5/deepmd",
	"./Y/dhcp/1-6/deepmd",
	"./Y/dhcp/1-7/deepmd",
	"./Y/dhcp/1-8/deepmd",
	"./Y/dhcp/1-9/deepmd",
	"./Y/dhcp/2-0/deepmd",
	"./Y/dhcp/2-1/deepmd",
	"./Y/dhcp/2-10/deepmd",
	"./Y/dhcp/2-11/deepmd",
	"./Y/dhcp/2-12/deepmd",
	"./Y/dhcp/2-13/deepmd",
	"./Y/dhcp/2-14/deepmd",
	"./Y/dhcp/2-15/deepmd",
	"./Y/dhcp/2-16/deepmd",
	"./Y/dhcp/2-17/deepmd",
	"./Y/dhcp/2-18/deepmd",
	"./Y/dhcp/2-19/deepmd",
	"./Y/dhcp/2-2/deepmd",
	"./Y/dhcp/2-3/deepmd",
	"./Y/dhcp/2-4/deepmd",
	"./Y/dhcp/2-5/deepmd",
	"./Y/dhcp/2-6/deepmd",
	"./Y/dhcp/2-7/deepmd",
	"./Y/dhcp/2-8/deepmd",
	"./Y/dhcp/2-9/deepmd",
	"./Y/dhcp/3-0/deepmd",
	"./Y/dhcp/3-1/deepmd",
	"./Y/dhcp/3-10/deepmd",
	"./Y/dhcp/3-11/deepmd",
	"./Y/dhcp/3-12/deepmd",
	"./Y/dhcp/3-13/deepmd",
	"./Y/dhcp/3-14/deepmd",
	"./Y/dhcp/3-15/deepmd",
	"./Y/dhcp/3-16/deepmd",
	"./Y/dhcp/3-17/deepmd",
	"./Y/dhcp/3-18/deepmd",
	"./Y/dhcp/3-19/deepmd",
	"./Y/dhcp/3-2/deepmd",
	"./Y/dhcp/3-3/deepmd",
	"./Y/dhcp/3-4/deepmd",
	"./Y/dhcp/3-5/deepmd",
	"./Y/dhcp/3-6/deepmd",
	"./Y/dhcp/3-7/deepmd",
	"./Y/dhcp/3-8/deepmd",
	"./Y/dhcp/3-9/deepmd",
"./Y/eos/mp-112/eos_00/task.000000/deepmd",
"./Y/eos/mp-112/eos_00/task.000001/deepmd",
"./Y/eos/mp-112/eos_00/task.000002/deepmd",
"./Y/eos/mp-112/eos_00/task.000003/deepmd",
"./Y/eos/mp-112/eos_00/task.000004/deepmd",
"./Y/eos/mp-112/eos_00/task.000005/deepmd",
"./Y/eos/mp-112/eos_00/task.000006/deepmd",
"./Y/eos/mp-112/eos_00/task.000007/deepmd",
"./Y/eos/mp-112/eos_00/task.000008/deepmd",
"./Y/eos/mp-112/eos_00/task.000009/deepmd",
"./Y/eos/mp-112/eos_00/task.000010/deepmd",
"./Y/eos/mp-112/eos_00/task.000011/deepmd",
"./Y/eos/mp-112/eos_00/task.000012/deepmd",
"./Y/eos/mp-112/eos_00/task.000013/deepmd",
"./Y/eos/mp-112/eos_00/task.000014/deepmd",
"./Y/eos/mp-112/eos_00/task.000015/deepmd",
"./Y/eos/mp-112/eos_00/task.000016/deepmd",
"./Y/eos/mp-112/eos_00/task.000017/deepmd",
"./Y/eos/mp-112/eos_00/task.000018/deepmd",
"./Y/eos/mp-112/eos_00/task.000019/deepmd",
"./Y/eos/mp-112/eos_00/task.000020/deepmd",
"./Y/eos/mp-112/eos_00/task.000021/deepmd",
"./Y/eos/mp-112/eos_00/task.000022/deepmd",
"./Y/eos/mp-112/eos_00/task.000023/deepmd",
"./Y/eos/mp-112/eos_00/task.000024/deepmd",
"./Y/eos/mp-112/eos_00/task.000025/deepmd",
"./Y/eos/mp-112/eos_00/task.000026/deepmd",
"./Y/eos/mp-112/eos_00/task.000027/deepmd",
"./Y/eos/mp-112/eos_00/task.000028/deepmd",
"./Y/eos/mp-112/eos_00/task.000029/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000000/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000001/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000002/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000003/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000004/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000005/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000006/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000007/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000008/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000009/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000010/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000011/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000012/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000013/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000014/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000015/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000016/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000017/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000018/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000019/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000020/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000021/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000022/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000023/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000024/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000025/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000026/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000027/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000028/deepmd",
"./Y/eos/mp-1187717/eos_00/task.000029/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000000/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000001/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000002/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000003/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000004/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000005/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000006/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000007/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000008/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000009/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000010/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000011/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000012/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000013/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000014/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000015/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000016/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000017/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000018/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000019/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000020/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000021/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000022/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000023/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000024/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000025/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000026/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000027/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000028/deepmd",
"./Y/eos/mp-1187739/eos_00/task.000029/deepmd",
"./Y/eos/mp-9/eos_00/task.000000/deepmd",
"./Y/eos/mp-9/eos_00/task.000001/deepmd",
"./Y/eos/mp-9/eos_00/task.000002/deepmd",
"./Y/eos/mp-9/eos_00/task.000003/deepmd",
"./Y/eos/mp-9/eos_00/task.000004/deepmd",
"./Y/eos/mp-9/eos_00/task.000005/deepmd",
"./Y/eos/mp-9/eos_00/task.000006/deepmd",
"./Y/eos/mp-9/eos_00/task.000007/deepmd",
"./Y/eos/mp-9/eos_00/task.000008/deepmd",
"./Y/eos/mp-9/eos_00/task.000009/deepmd",
"./Y/eos/mp-9/eos_00/task.000010/deepmd",
"./Y/eos/mp-9/eos_00/task.000011/deepmd",
"./Y/eos/mp-9/eos_00/task.000012/deepmd",
"./Y/eos/mp-9/eos_00/task.000013/deepmd",
"./Y/eos/mp-9/eos_00/task.000014/deepmd",
"./Y/eos/mp-9/eos_00/task.000015/deepmd",
"./Y/eos/mp-9/eos_00/task.000016/deepmd",
"./Y/eos/mp-9/eos_00/task.000017/deepmd",
"./Y/eos/mp-9/eos_00/task.000018/deepmd",
"./Y/eos/mp-9/eos_00/task.000019/deepmd",
"./Y/eos/mp-9/eos_00/task.000020/deepmd",
"./Y/eos/mp-9/eos_00/task.000021/deepmd",
"./Y/eos/mp-9/eos_00/task.000022/deepmd",
"./Y/eos/mp-9/eos_00/task.000023/deepmd",
"./Y/eos/mp-9/eos_00/task.000024/deepmd",
"./Y/eos/mp-9/eos_00/task.000025/deepmd",
"./Y/eos/mp-9/eos_00/task.000026/deepmd",
"./Y/eos/mp-9/eos_00/task.000027/deepmd",
"./Y/eos/mp-9/eos_00/task.000028/deepmd",
"./Y/eos/mp-9/eos_00/task.000029/deepmd",
"./Y/init-new/9R/POSCAR_1187739.02x02x01/02.md/sys-0036/deepmd",
"./Y/init-new/dhcp/POSCAR_1187717.02x02x01/02.md/sys-0016/deepmd",
"./Y/init-new/fcc/POSCAR_9.02x02x02/02.md/sys-0032/deepmd",
"./Y/init-new/hcp/POSCAR_112.02x02x02/02.md/sys-0016/deepmd",

"./Y/relable/iter.000010/02.fp/task.002.000000/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000001/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000002/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000003/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000004/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000005/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000006/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000007/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000008/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000009/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000010/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000011/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000012/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000013/deepmd",
"./Y/relable/iter.000010/02.fp/task.002.000014/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000000/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000001/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000002/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000003/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000004/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000005/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000006/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000007/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000008/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000009/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000010/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000011/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000012/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000013/deepmd",
"./Y/relable/iter.000010/02.fp/task.010.000014/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000000/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000001/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000002/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000003/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000004/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000005/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000006/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000007/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000008/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000009/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000010/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000011/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000012/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000013/deepmd",
"./Y/relable/iter.000010/02.fp/task.018.000014/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000000/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000001/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000002/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000003/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000004/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000005/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000006/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000007/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000008/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000009/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000010/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000011/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000012/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000013/deepmd",
"./Y/relable/iter.000011/02.fp/task.003.000014/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000000/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000001/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000002/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000003/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000004/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000005/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000006/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000007/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000008/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000009/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000010/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000011/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000012/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000013/deepmd",
"./Y/relable/iter.000011/02.fp/task.011.000014/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000000/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000001/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000002/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000003/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000004/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000005/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000006/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000007/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000008/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000009/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000010/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000011/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000012/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000013/deepmd",
"./Y/relable/iter.000011/02.fp/task.019.000014/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000000/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000001/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000002/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000003/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000004/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000005/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000006/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000007/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000008/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000009/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000010/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000011/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000012/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000013/deepmd",
"./Y/relable/iter.000012/02.fp/task.004.000014/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000000/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000001/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000002/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000003/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000004/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000005/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000006/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000007/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000008/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000009/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000010/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000011/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000012/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000013/deepmd",
"./Y/relable/iter.000012/02.fp/task.012.000014/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000000/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000001/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000002/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000003/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000004/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000005/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000006/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000007/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000008/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000009/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000010/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000011/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000012/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000013/deepmd",
"./Y/relable/iter.000012/02.fp/task.020.000014/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000000/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000001/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000002/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000003/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000004/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000005/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000006/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000007/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000008/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000009/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000010/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000011/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000012/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000013/deepmd",
"./Y/relable/iter.000013/02.fp/task.005.000014/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000000/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000001/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000002/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000003/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000004/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000005/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000006/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000007/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000008/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000009/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000010/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000011/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000012/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000013/deepmd",
"./Y/relable/iter.000013/02.fp/task.013.000014/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000000/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000001/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000002/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000003/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000004/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000005/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000006/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000007/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000008/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000009/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000010/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000011/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000012/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000013/deepmd",
"./Y/relable/iter.000013/02.fp/task.021.000014/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000000/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000001/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000002/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000003/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000004/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000005/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000006/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000007/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000008/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000009/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000010/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000011/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000012/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000013/deepmd",
"./Y/relable/iter.000014/02.fp/task.006.000014/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000000/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000001/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000002/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000003/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000004/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000005/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000006/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000007/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000008/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000009/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000010/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000011/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000012/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000013/deepmd",
"./Y/relable/iter.000014/02.fp/task.014.000014/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000000/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000001/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000002/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000003/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000004/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000005/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000006/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000007/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000008/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000009/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000010/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000011/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000012/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000013/deepmd",
"./Y/relable/iter.000014/02.fp/task.022.000014/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000000/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000001/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000002/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000003/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000004/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000005/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000006/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000007/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000008/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000009/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000010/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000011/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000012/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000013/deepmd",
"./Y/relable/iter.000015/02.fp/task.007.000014/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000000/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000001/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000002/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000003/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000004/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000005/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000006/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000007/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000008/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000009/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000010/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000011/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000012/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000013/deepmd",
"./Y/relable/iter.000015/02.fp/task.015.000014/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000000/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000001/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000002/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000003/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000004/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000005/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000006/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000007/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000008/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000009/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000010/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000011/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000012/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000013/deepmd",
"./Y/relable/iter.000015/02.fp/task.023.000014/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000000/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000001/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000002/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000003/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000004/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000005/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000006/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000007/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000008/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000009/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000010/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000011/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000012/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000013/deepmd",
"./Y/relable/iter.000016/02.fp/task.000.000014/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000000/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000001/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000002/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000003/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000004/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000005/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000006/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000007/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000008/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000009/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000010/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000011/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000012/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000013/deepmd",
"./Y/relable/iter.000016/02.fp/task.008.000014/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000000/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000001/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000002/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000003/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000004/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000005/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000006/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000007/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000008/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000009/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000010/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000011/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000012/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000013/deepmd",
"./Y/relable/iter.000016/02.fp/task.016.000014/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000000/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000001/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000002/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000003/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000004/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000005/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000006/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000007/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000008/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000009/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000010/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000011/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000012/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000013/deepmd",
"./Y/relable/iter.000017/02.fp/task.001.000014/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000000/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000001/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000002/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000003/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000004/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000005/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000006/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000007/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000008/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000009/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000010/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000011/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000012/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000013/deepmd",
"./Y/relable/iter.000017/02.fp/task.009.000014/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000000/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000001/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000002/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000003/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000004/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000005/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000006/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000008/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000009/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000010/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000011/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000012/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000013/deepmd",
"./Y/relable/iter.000017/02.fp/task.017.000014/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000000/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000001/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000002/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000003/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000004/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000005/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000006/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000007/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000008/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000009/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000010/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000011/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000012/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000013/deepmd",
"./Y/relable/iter.000018/02.fp/task.002.000014/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000000/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000001/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000002/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000003/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000004/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000005/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000006/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000007/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000008/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000009/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000010/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000011/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000012/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000013/deepmd",
"./Y/relable/iter.000018/02.fp/task.010.000014/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000000/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000001/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000002/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000003/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000004/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000005/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000006/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000007/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000008/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000009/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000010/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000011/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000012/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000013/deepmd",
"./Y/relable/iter.000018/02.fp/task.018.000014/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000000/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000001/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000002/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000003/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000004/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000005/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000006/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000007/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000008/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000009/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000010/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000011/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000012/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000013/deepmd",
"./Y/relable/iter.000019/02.fp/task.003.000014/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000000/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000001/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000002/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000003/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000004/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000005/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000006/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000007/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000008/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000009/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000010/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000011/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000012/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000013/deepmd",
"./Y/relable/iter.000019/02.fp/task.011.000014/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000000/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000001/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000002/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000003/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000004/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000005/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000006/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000007/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000008/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000009/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000010/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000011/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000012/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000013/deepmd",
"./Y/relable/iter.000019/02.fp/task.019.000014/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000000/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000001/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000002/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000003/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000004/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000005/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000006/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000007/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000008/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000009/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000010/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000011/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000012/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000013/deepmd",
"./Y/relable/iter.000020/02.fp/task.004.000014/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000000/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000001/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000002/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000003/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000004/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000005/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000007/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000008/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000009/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000010/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000011/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000012/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000013/deepmd",
"./Y/relable/iter.000020/02.fp/task.012.000014/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000000/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000002/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000003/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000004/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000005/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000006/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000007/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000008/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000009/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000010/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000011/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000012/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000013/deepmd",
"./Y/relable/iter.000020/02.fp/task.020.000014/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000000/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000001/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000002/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000003/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000004/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000005/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000007/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000008/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000009/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000010/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000011/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000012/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000013/deepmd",
"./Y/relable/iter.000021/02.fp/task.005.000014/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000000/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000001/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000002/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000003/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000004/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000005/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000006/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000007/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000008/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000009/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000010/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000011/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000012/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000013/deepmd",
"./Y/relable/iter.000021/02.fp/task.013.000014/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000000/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000001/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000002/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000003/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000004/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000005/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000006/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000007/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000008/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000009/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000010/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000011/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000012/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000013/deepmd",
"./Y/relable/iter.000021/02.fp/task.021.000014/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000000/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000001/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000002/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000003/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000004/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000005/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000006/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000007/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000008/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000009/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000010/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000011/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000012/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000013/deepmd",
"./Y/relable/iter.000022/02.fp/task.006.000014/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000000/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000001/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000002/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000003/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000004/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000005/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000006/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000007/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000008/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000009/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000010/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000011/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000012/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000013/deepmd",
"./Y/relable/iter.000022/02.fp/task.014.000014/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000000/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000001/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000002/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000003/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000004/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000005/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000006/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000007/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000008/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000009/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000010/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000011/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000012/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000013/deepmd",
"./Y/relable/iter.000022/02.fp/task.022.000014/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000000/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000001/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000002/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000003/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000004/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000005/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000006/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000007/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000008/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000009/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000010/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000011/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000012/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000013/deepmd",
"./Y/relable/iter.000023/02.fp/task.007.000014/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000000/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000001/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000002/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000003/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000004/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000005/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000006/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000007/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000008/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000009/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000010/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000011/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000012/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000013/deepmd",
"./Y/relable/iter.000023/02.fp/task.015.000014/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000000/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000001/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000002/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000003/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000004/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000005/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000006/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000007/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000008/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000009/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000010/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000011/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000012/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000013/deepmd",
"./Y/relable/iter.000023/02.fp/task.023.000014/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000000/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000001/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000002/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000003/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000004/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000005/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000006/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000007/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000008/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000009/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000010/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000011/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000012/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000013/deepmd",
"./Y/relable/iter.000024/02.fp/task.000.000014/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000000/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000001/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000002/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000003/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000004/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000005/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000006/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000007/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000008/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000009/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000010/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000011/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000012/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000013/deepmd",
"./Y/relable/iter.000024/02.fp/task.008.000014/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000000/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000001/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000002/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000003/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000004/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000005/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000006/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000007/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000008/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000009/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000010/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000011/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000012/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000013/deepmd",
"./Y/relable/iter.000024/02.fp/task.016.000014/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000000/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000001/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000002/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000003/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000004/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000005/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000006/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000007/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000008/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000009/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000010/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000011/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000012/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000013/deepmd",
"./Y/relable/iter.000025/02.fp/task.001.000014/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000000/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000001/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000002/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000003/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000004/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000005/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000006/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000007/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000008/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000009/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000010/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000011/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000012/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000013/deepmd",
"./Y/relable/iter.000025/02.fp/task.009.000014/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000000/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000001/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000002/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000003/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000004/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000005/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000006/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000007/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000008/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000009/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000010/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000011/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000012/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000013/deepmd",
"./Y/relable/iter.000025/02.fp/task.017.000014/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000000/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000001/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000002/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000003/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000004/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000005/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000006/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000007/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000008/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000009/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000010/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000011/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000012/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000013/deepmd",
"./Y/relable/iter.000026/02.fp/task.002.000014/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000000/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000001/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000002/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000003/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000004/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000005/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000006/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000007/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000008/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000009/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000010/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000011/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000012/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000013/deepmd",
"./Y/relable/iter.000026/02.fp/task.010.000014/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000000/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000001/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000002/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000003/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000004/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000005/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000006/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000007/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000008/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000009/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000010/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000011/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000012/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000013/deepmd",
"./Y/relable/iter.000026/02.fp/task.018.000014/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000000/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000001/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000002/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000003/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000004/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000005/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000006/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000007/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000008/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000009/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000010/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000011/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000012/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000013/deepmd",
"./Y/relable/iter.000027/02.fp/task.003.000014/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000000/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000001/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000002/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000003/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000004/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000005/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000006/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000008/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000009/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000010/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000011/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000012/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000013/deepmd",
"./Y/relable/iter.000027/02.fp/task.011.000014/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000000/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000001/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000002/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000003/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000004/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000005/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000006/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000007/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000008/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000009/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000010/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000011/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000012/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000013/deepmd",
"./Y/relable/iter.000027/02.fp/task.019.000014/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000000/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000002/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000003/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000004/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000005/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000006/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000007/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000008/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000009/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000011/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000012/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000013/deepmd",
"./Y/relable/iter.000028/02.fp/task.004.000014/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000000/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000002/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000003/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000004/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000005/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000006/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000007/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000008/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000009/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000011/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000012/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000013/deepmd",
"./Y/relable/iter.000028/02.fp/task.012.000014/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000000/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000001/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000002/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000003/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000004/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000005/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000006/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000007/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000008/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000009/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000010/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000011/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000012/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000013/deepmd",
"./Y/relable/iter.000028/02.fp/task.020.000014/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000000/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000001/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000002/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000003/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000005/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000006/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000007/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000008/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000009/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000010/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000011/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000012/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000013/deepmd",
"./Y/relable/iter.000029/02.fp/task.005.000014/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000000/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000001/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000002/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000003/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000004/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000005/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000006/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000007/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000008/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000009/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000010/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000011/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000012/deepmd",
"./Y/relable/iter.000029/02.fp/task.013.000013/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000000/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000001/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000002/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000003/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000004/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000005/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000006/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000007/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000008/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000009/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000010/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000012/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000013/deepmd",
"./Y/relable/iter.000029/02.fp/task.021.000014/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000000/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000001/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000002/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000003/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000004/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000005/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000006/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000007/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000008/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000009/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000010/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000011/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000012/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000013/deepmd",
"./Y/relable/iter.000030/02.fp/task.006.000014/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000000/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000001/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000002/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000003/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000004/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000005/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000006/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000007/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000008/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000009/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000010/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000011/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000012/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000013/deepmd",
"./Y/relable/iter.000030/02.fp/task.014.000014/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000000/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000002/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000003/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000004/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000005/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000006/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000007/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000008/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000009/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000010/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000011/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000012/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000013/deepmd",
"./Y/relable/iter.000030/02.fp/task.022.000014/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000000/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000001/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000002/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000003/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000004/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000005/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000006/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000007/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000008/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000009/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000010/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000011/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000012/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000013/deepmd",
"./Y/relable/iter.000031/02.fp/task.007.000014/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000000/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000001/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000002/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000003/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000004/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000005/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000006/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000007/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000008/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000009/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000010/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000011/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000012/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000013/deepmd",
"./Y/relable/iter.000031/02.fp/task.015.000014/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000000/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000001/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000002/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000003/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000004/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000005/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000006/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000007/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000008/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000009/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000010/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000011/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000012/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000013/deepmd",
"./Y/relable/iter.000031/02.fp/task.023.000014/deepmd",

"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR0/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR1/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR10/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR11/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR12/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR13/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR14/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR2/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR3/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR4/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR5/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR6/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR7/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR8/deepmd",
"./Y/relable/iter.000000/02.fp/data.000/aPOSCAR9/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR0/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR1/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR10/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR11/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR12/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR13/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR14/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR2/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR3/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR4/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR5/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR6/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR7/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR8/deepmd",
"./Y/relable/iter.000000/02.fp/data.001/aPOSCAR9/deepmd",
"./Y/relable/iter.000000/02.fp/data.002/aPOSCAR0/deepmd",
"./Y/relable/iter.000000/02.fp/data.002/aPOSCAR1/deepmd",
"./Y/relable/iter.000000/02.fp/data.002/aPOSCAR10/deepmd",
"./Y/relable/iter.000000/02.fp/data.002/aPOSCAR2/deepmd",
"./Y/relable/iter.000000/02.fp/data.002/aPOSCAR3/deepmd",
"./Y/relable/iter.000000/02.fp/data.002/aPOSCAR4/deepmd",
"./Y/relable/iter.000000/02.fp/data.002/aPOSCAR5/deepmd",
"./Y/relable/iter.000000/02.fp/data.002/aPOSCAR6/deepmd",
"./Y/relable/iter.000000/02.fp/data.002/aPOSCAR7/deepmd",
"./Y/relable/iter.000000/02.fp/data.002/aPOSCAR8/deepmd",
"./Y/relable/iter.000000/02.fp/data.002/aPOSCAR9/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR0/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR1/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR10/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR11/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR12/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR13/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR14/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR2/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR3/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR4/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR5/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR6/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR7/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR8/deepmd",
"./Y/relable/iter.000000/02.fp/data.003/aPOSCAR9/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR0/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR1/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR10/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR11/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR12/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR13/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR14/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR2/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR3/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR4/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR5/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR6/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR7/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR8/deepmd",
"./Y/relable/iter.000000/02.fp/data.008/aPOSCAR9/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR0/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR1/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR10/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR11/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR12/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR13/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR14/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR2/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR3/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR4/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR5/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR6/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR7/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR8/deepmd",
"./Y/relable/iter.000000/02.fp/data.009/aPOSCAR9/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR0/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR1/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR10/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR11/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR12/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR13/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR14/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR2/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR3/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR4/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR5/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR6/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR7/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR8/deepmd",
"./Y/relable/iter.000000/02.fp/data.010/aPOSCAR9/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR0/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR1/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR10/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR11/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR12/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR13/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR14/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR2/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR3/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR4/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR5/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR6/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR7/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR8/deepmd",
"./Y/relable/iter.000000/02.fp/data.011/aPOSCAR9/deepmd",
"./Y/relable/iter.000000/02.fp/data.016/aPOSCAR0/deepmd",
"./Y/relable/iter.000000/02.fp/data.016/aPOSCAR1/deepmd",
"./Y/relable/iter.000000/02.fp/data.016/aPOSCAR10/deepmd",
"./Y/relable/iter.000000/02.fp/data.016/aPOSCAR2/deepmd",
"./Y/relable/iter.000000/02.fp/data.016/aPOSCAR3/deepmd",
"./Y/relable/iter.000000/02.fp/data.016/aPOSCAR4/deepmd",
"./Y/relable/iter.000000/02.fp/data.016/aPOSCAR5/deepmd",
"./Y/relable/iter.000000/02.fp/data.016/aPOSCAR6/deepmd",
"./Y/relable/iter.000000/02.fp/data.016/aPOSCAR7/deepmd",
"./Y/relable/iter.000000/02.fp/data.016/aPOSCAR8/deepmd",
"./Y/relable/iter.000000/02.fp/data.016/aPOSCAR9/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR0/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR1/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR10/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR11/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR12/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR13/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR14/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR2/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR3/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR4/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR5/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR6/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR7/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR8/deepmd",
"./Y/relable/iter.000000/02.fp/data.017/aPOSCAR9/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR0/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR1/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR10/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR11/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR12/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR13/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR14/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR2/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR3/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR4/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR5/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR6/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR7/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR8/deepmd",
"./Y/relable/iter.000000/02.fp/data.018/aPOSCAR9/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR0/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR1/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR10/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR11/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR12/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR13/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR14/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR2/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR3/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR4/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR5/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR6/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR7/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR8/deepmd",
"./Y/relable/iter.000000/02.fp/data.019/aPOSCAR9/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR0/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR1/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR10/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR11/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR12/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR13/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR14/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR2/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR3/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR4/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR5/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR6/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR7/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR8/deepmd",
"./Y/relable/iter.000001/02.fp/data.004/aPOSCAR9/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR0/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR1/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR10/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR11/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR12/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR13/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR14/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR2/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR3/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR4/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR5/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR6/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR7/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR8/deepmd",
"./Y/relable/iter.000001/02.fp/data.005/aPOSCAR9/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR0/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR1/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR10/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR11/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR12/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR13/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR14/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR2/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR3/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR4/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR5/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR6/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR7/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR8/deepmd",
"./Y/relable/iter.000001/02.fp/data.006/aPOSCAR9/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR0/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR1/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR10/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR11/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR12/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR2/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR3/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR4/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR5/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR6/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR7/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR8/deepmd",
"./Y/relable/iter.000001/02.fp/data.007/aPOSCAR9/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR0/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR1/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR10/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR11/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR12/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR2/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR3/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR4/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR5/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR6/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR7/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR8/deepmd",
"./Y/relable/iter.000001/02.fp/data.012/aPOSCAR9/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR0/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR1/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR10/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR11/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR12/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR13/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR14/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR2/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR3/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR4/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR5/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR6/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR7/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR8/deepmd",
"./Y/relable/iter.000001/02.fp/data.013/aPOSCAR9/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR0/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR1/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR10/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR11/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR12/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR13/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR14/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR2/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR3/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR4/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR5/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR6/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR7/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR8/deepmd",
"./Y/relable/iter.000001/02.fp/data.014/aPOSCAR9/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR0/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR1/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR10/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR11/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR12/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR13/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR14/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR2/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR3/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR4/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR5/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR6/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR7/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR8/deepmd",
"./Y/relable/iter.000001/02.fp/data.015/aPOSCAR9/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR0/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR1/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR10/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR11/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR12/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR13/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR14/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR2/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR3/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR4/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR5/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR6/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR7/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR8/deepmd",
"./Y/relable/iter.000001/02.fp/data.020/aPOSCAR9/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR0/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR1/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR10/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR11/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR12/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR2/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR3/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR4/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR5/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR6/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR7/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR8/deepmd",
"./Y/relable/iter.000001/02.fp/data.021/aPOSCAR9/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR0/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR1/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR10/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR11/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR12/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR2/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR3/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR4/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR5/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR6/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR7/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR8/deepmd",
"./Y/relable/iter.000001/02.fp/data.022/aPOSCAR9/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR0/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR1/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR10/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR11/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR12/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR13/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR14/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR2/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR3/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR4/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR5/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR6/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR7/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR8/deepmd",
"./Y/relable/iter.000001/02.fp/data.023/aPOSCAR9/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR0/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR1/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR10/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR11/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR12/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR13/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR14/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR2/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR3/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR4/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR5/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR6/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR7/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR8/deepmd",
"./Y/relable/iter.000002/02.fp/data.000/aPOSCAR9/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR0/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR1/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR10/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR11/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR12/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR13/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR14/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR2/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR3/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR4/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR5/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR6/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR7/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR8/deepmd",
"./Y/relable/iter.000002/02.fp/data.001/aPOSCAR9/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR0/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR1/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR10/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR11/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR12/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR13/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR14/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR2/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR3/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR4/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR5/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR6/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR7/deepmd",
"./Y/relable/iter.000002/02.fp/data.002/aPOSCAR8/deepmd",
"./Y/relable/iter.000002/02.fp/data.003/aPOSCAR0/deepmd",
"./Y/relable/iter.000002/02.fp/data.003/aPOSCAR1/deepmd",
"./Y/relable/iter.000002/02.fp/data.003/aPOSCAR10/deepmd",
"./Y/relable/iter.000002/02.fp/data.003/aPOSCAR2/deepmd",
"./Y/relable/iter.000002/02.fp/data.003/aPOSCAR3/deepmd",
"./Y/relable/iter.000002/02.fp/data.003/aPOSCAR4/deepmd",
"./Y/relable/iter.000002/02.fp/data.003/aPOSCAR5/deepmd",
"./Y/relable/iter.000002/02.fp/data.003/aPOSCAR6/deepmd",
"./Y/relable/iter.000002/02.fp/data.003/aPOSCAR7/deepmd",
"./Y/relable/iter.000002/02.fp/data.003/aPOSCAR8/deepmd",
"./Y/relable/iter.000002/02.fp/data.003/aPOSCAR9/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR0/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR1/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR10/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR11/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR12/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR13/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR14/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR2/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR3/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR4/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR5/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR6/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR7/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR8/deepmd",
"./Y/relable/iter.000002/02.fp/data.008/aPOSCAR9/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR0/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR1/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR10/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR11/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR12/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR13/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR14/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR2/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR3/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR4/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR5/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR6/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR7/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR8/deepmd",
"./Y/relable/iter.000002/02.fp/data.009/aPOSCAR9/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR0/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR1/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR10/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR11/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR12/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR13/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR14/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR2/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR3/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR4/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR5/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR6/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR7/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR8/deepmd",
"./Y/relable/iter.000002/02.fp/data.010/aPOSCAR9/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR0/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR1/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR10/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR11/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR12/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR13/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR14/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR2/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR3/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR4/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR5/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR6/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR7/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR8/deepmd",
"./Y/relable/iter.000002/02.fp/data.011/aPOSCAR9/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR0/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR1/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR10/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR11/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR12/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR13/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR14/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR2/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR3/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR4/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR5/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR6/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR7/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR8/deepmd",
"./Y/relable/iter.000002/02.fp/data.016/aPOSCAR9/deepmd",
"./Y/relable/iter.000002/02.fp/data.017/aPOSCAR0/deepmd",
"./Y/relable/iter.000002/02.fp/data.017/aPOSCAR1/deepmd",
"./Y/relable/iter.000002/02.fp/data.017/aPOSCAR10/deepmd",
"./Y/relable/iter.000002/02.fp/data.017/aPOSCAR2/deepmd",
"./Y/relable/iter.000002/02.fp/data.017/aPOSCAR3/deepmd",
"./Y/relable/iter.000002/02.fp/data.017/aPOSCAR4/deepmd",
"./Y/relable/iter.000002/02.fp/data.017/aPOSCAR5/deepmd",
"./Y/relable/iter.000002/02.fp/data.017/aPOSCAR6/deepmd",
"./Y/relable/iter.000002/02.fp/data.017/aPOSCAR7/deepmd",
"./Y/relable/iter.000002/02.fp/data.017/aPOSCAR8/deepmd",
"./Y/relable/iter.000002/02.fp/data.017/aPOSCAR9/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR0/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR1/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR10/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR11/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR12/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR13/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR14/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR2/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR3/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR4/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR5/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR6/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR7/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR8/deepmd",
"./Y/relable/iter.000002/02.fp/data.018/aPOSCAR9/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR0/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR1/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR10/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR11/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR12/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR13/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR14/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR2/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR3/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR4/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR5/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR6/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR7/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR8/deepmd",
"./Y/relable/iter.000002/02.fp/data.019/aPOSCAR9/deepmd",
"./Y/relable/iter.000003/02.fp/data.004/aPOSCAR0/deepmd",
"./Y/relable/iter.000003/02.fp/data.004/aPOSCAR1/deepmd",
"./Y/relable/iter.000003/02.fp/data.004/aPOSCAR2/deepmd",
"./Y/relable/iter.000003/02.fp/data.004/aPOSCAR3/deepmd",
"./Y/relable/iter.000003/02.fp/data.004/aPOSCAR4/deepmd",
"./Y/relable/iter.000003/02.fp/data.005/aPOSCAR0/deepmd",
"./Y/relable/iter.000003/02.fp/data.005/aPOSCAR1/deepmd",
"./Y/relable/iter.000003/02.fp/data.005/aPOSCAR2/deepmd",
"./Y/relable/iter.000003/02.fp/data.005/aPOSCAR3/deepmd",
"./Y/relable/iter.000003/02.fp/data.005/aPOSCAR4/deepmd",
"./Y/relable/iter.000003/02.fp/data.005/aPOSCAR5/deepmd",
"./Y/relable/iter.000003/02.fp/data.006/aPOSCAR0/deepmd",
"./Y/relable/iter.000003/02.fp/data.006/aPOSCAR1/deepmd",
"./Y/relable/iter.000003/02.fp/data.006/aPOSCAR2/deepmd",
"./Y/relable/iter.000003/02.fp/data.006/aPOSCAR3/deepmd",
"./Y/relable/iter.000003/02.fp/data.006/aPOSCAR4/deepmd",
"./Y/relable/iter.000003/02.fp/data.007/aPOSCAR0/deepmd",
"./Y/relable/iter.000003/02.fp/data.007/aPOSCAR1/deepmd",
"./Y/relable/iter.000003/02.fp/data.007/aPOSCAR2/deepmd",
"./Y/relable/iter.000003/02.fp/data.007/aPOSCAR3/deepmd",
"./Y/relable/iter.000003/02.fp/data.007/aPOSCAR4/deepmd",
"./Y/relable/iter.000003/02.fp/data.007/aPOSCAR5/deepmd",
"./Y/relable/iter.000003/02.fp/data.007/aPOSCAR6/deepmd",
"./Y/relable/iter.000003/02.fp/data.007/aPOSCAR7/deepmd",
"./Y/relable/iter.000003/02.fp/data.007/aPOSCAR8/deepmd",
"./Y/relable/iter.000003/02.fp/data.007/aPOSCAR9/deepmd",
"./Y/relable/iter.000003/02.fp/data.012/aPOSCAR0/deepmd",
"./Y/relable/iter.000003/02.fp/data.012/aPOSCAR1/deepmd",
"./Y/relable/iter.000003/02.fp/data.012/aPOSCAR2/deepmd",
"./Y/relable/iter.000003/02.fp/data.012/aPOSCAR3/deepmd",
"./Y/relable/iter.000003/02.fp/data.012/aPOSCAR4/deepmd",
"./Y/relable/iter.000003/02.fp/data.012/aPOSCAR5/deepmd",
"./Y/relable/iter.000003/02.fp/data.012/aPOSCAR6/deepmd",
"./Y/relable/iter.000003/02.fp/data.012/aPOSCAR7/deepmd",
"./Y/relable/iter.000003/02.fp/data.012/aPOSCAR8/deepmd",
"./Y/relable/iter.000003/02.fp/data.012/aPOSCAR9/deepmd",
"./Y/relable/iter.000003/02.fp/data.013/aPOSCAR0/deepmd",
"./Y/relable/iter.000003/02.fp/data.013/aPOSCAR1/deepmd",
"./Y/relable/iter.000003/02.fp/data.013/aPOSCAR2/deepmd",
"./Y/relable/iter.000003/02.fp/data.013/aPOSCAR3/deepmd",
"./Y/relable/iter.000003/02.fp/data.013/aPOSCAR4/deepmd",
"./Y/relable/iter.000003/02.fp/data.013/aPOSCAR5/deepmd",
"./Y/relable/iter.000003/02.fp/data.013/aPOSCAR6/deepmd",
"./Y/relable/iter.000003/02.fp/data.013/aPOSCAR7/deepmd",
"./Y/relable/iter.000003/02.fp/data.013/aPOSCAR8/deepmd",
"./Y/relable/iter.000003/02.fp/data.014/aPOSCAR0/deepmd",
"./Y/relable/iter.000003/02.fp/data.014/aPOSCAR1/deepmd",
"./Y/relable/iter.000003/02.fp/data.014/aPOSCAR2/deepmd",
"./Y/relable/iter.000003/02.fp/data.014/aPOSCAR3/deepmd",
"./Y/relable/iter.000003/02.fp/data.014/aPOSCAR4/deepmd",
"./Y/relable/iter.000003/02.fp/data.014/aPOSCAR5/deepmd",
"./Y/relable/iter.000003/02.fp/data.014/aPOSCAR6/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR0/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR1/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR10/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR11/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR12/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR13/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR2/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR3/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR4/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR5/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR6/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR7/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR8/deepmd",
"./Y/relable/iter.000003/02.fp/data.015/aPOSCAR9/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR0/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR1/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR10/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR11/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR12/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR13/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR14/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR3/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR4/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR5/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR6/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR7/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR8/deepmd",
"./Y/relable/iter.000003/02.fp/data.020/aPOSCAR9/deepmd",
"./Y/relable/iter.000003/02.fp/data.021/aPOSCAR0/deepmd",
"./Y/relable/iter.000003/02.fp/data.021/aPOSCAR1/deepmd",
"./Y/relable/iter.000003/02.fp/data.021/aPOSCAR10/deepmd",
"./Y/relable/iter.000003/02.fp/data.021/aPOSCAR11/deepmd",
"./Y/relable/iter.000003/02.fp/data.021/aPOSCAR2/deepmd",
"./Y/relable/iter.000003/02.fp/data.021/aPOSCAR3/deepmd",
"./Y/relable/iter.000003/02.fp/data.021/aPOSCAR4/deepmd",
"./Y/relable/iter.000003/02.fp/data.021/aPOSCAR5/deepmd",
"./Y/relable/iter.000003/02.fp/data.021/aPOSCAR6/deepmd",
"./Y/relable/iter.000003/02.fp/data.021/aPOSCAR7/deepmd",
"./Y/relable/iter.000003/02.fp/data.021/aPOSCAR8/deepmd",
"./Y/relable/iter.000003/02.fp/data.021/aPOSCAR9/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR0/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR1/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR10/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR11/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR12/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR13/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR14/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR2/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR3/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR4/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR5/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR6/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR7/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR8/deepmd",
"./Y/relable/iter.000003/02.fp/data.022/aPOSCAR9/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR0/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR1/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR10/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR11/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR12/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR13/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR2/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR3/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR4/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR5/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR6/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR7/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR8/deepmd",
"./Y/relable/iter.000003/02.fp/data.023/aPOSCAR9/deepmd",
"./Y/relable/iter.000006/02.fp/data.000/aPOSCAR0/deepmd",
"./Y/relable/iter.000006/02.fp/data.000/aPOSCAR1/deepmd",
"./Y/relable/iter.000006/02.fp/data.000/aPOSCAR2/deepmd",
"./Y/relable/iter.000006/02.fp/data.000/aPOSCAR3/deepmd",
"./Y/relable/iter.000006/02.fp/data.001/aPOSCAR0/deepmd",
"./Y/relable/iter.000006/02.fp/data.001/aPOSCAR1/deepmd",
"./Y/relable/iter.000006/02.fp/data.001/aPOSCAR2/deepmd",
"./Y/relable/iter.000006/02.fp/data.001/aPOSCAR3/deepmd",
"./Y/relable/iter.000006/02.fp/data.002/aPOSCAR0/deepmd",
"./Y/relable/iter.000006/02.fp/data.002/aPOSCAR1/deepmd",
"./Y/relable/iter.000006/02.fp/data.003/aPOSCAR0/deepmd",
"./Y/relable/iter.000006/02.fp/data.003/aPOSCAR1/deepmd",
"./Y/relable/iter.000006/02.fp/data.003/aPOSCAR2/deepmd",
"./Y/relable/iter.000006/02.fp/data.003/aPOSCAR3/deepmd",
"./Y/relable/iter.000006/02.fp/data.003/aPOSCAR4/deepmd",
"./Y/relable/iter.000006/02.fp/data.008/aPOSCAR0/deepmd",
"./Y/relable/iter.000006/02.fp/data.008/aPOSCAR1/deepmd",
"./Y/relable/iter.000006/02.fp/data.008/aPOSCAR2/deepmd",
"./Y/relable/iter.000006/02.fp/data.008/aPOSCAR3/deepmd",
"./Y/relable/iter.000006/02.fp/data.008/aPOSCAR4/deepmd",
"./Y/relable/iter.000006/02.fp/data.008/aPOSCAR5/deepmd",
"./Y/relable/iter.000006/02.fp/data.009/aPOSCAR0/deepmd",
"./Y/relable/iter.000006/02.fp/data.009/aPOSCAR1/deepmd",
"./Y/relable/iter.000006/02.fp/data.009/aPOSCAR2/deepmd",
"./Y/relable/iter.000006/02.fp/data.009/aPOSCAR3/deepmd",
"./Y/relable/iter.000006/02.fp/data.009/aPOSCAR4/deepmd",
"./Y/relable/iter.000006/02.fp/data.010/aPOSCAR0/deepmd",
"./Y/relable/iter.000006/02.fp/data.010/aPOSCAR1/deepmd",
"./Y/relable/iter.000006/02.fp/data.010/aPOSCAR2/deepmd",
"./Y/relable/iter.000006/02.fp/data.010/aPOSCAR3/deepmd",
"./Y/relable/iter.000006/02.fp/data.011/aPOSCAR0/deepmd",
"./Y/relable/iter.000006/02.fp/data.011/aPOSCAR1/deepmd",
"./Y/relable/iter.000006/02.fp/data.011/aPOSCAR2/deepmd",
"./Y/relable/iter.000006/02.fp/data.011/aPOSCAR3/deepmd",
"./Y/relable/iter.000006/02.fp/data.011/aPOSCAR4/deepmd",
"./Y/relable/iter.000006/02.fp/data.011/aPOSCAR5/deepmd",
"./Y/relable/iter.000006/02.fp/data.011/aPOSCAR6/deepmd",
"./Y/relable/iter.000006/02.fp/data.011/aPOSCAR7/deepmd",
"./Y/relable/iter.000006/02.fp/data.011/aPOSCAR8/deepmd",
"./Y/relable/iter.000006/02.fp/data.016/aPOSCAR0/deepmd",
"./Y/relable/iter.000006/02.fp/data.016/aPOSCAR1/deepmd",
"./Y/relable/iter.000006/02.fp/data.016/aPOSCAR2/deepmd",
"./Y/relable/iter.000006/02.fp/data.016/aPOSCAR3/deepmd",
"./Y/relable/iter.000006/02.fp/data.016/aPOSCAR4/deepmd",
"./Y/relable/iter.000006/02.fp/data.016/aPOSCAR5/deepmd",
"./Y/relable/iter.000006/02.fp/data.016/aPOSCAR6/deepmd",
"./Y/relable/iter.000006/02.fp/data.017/aPOSCAR0/deepmd",
"./Y/relable/iter.000006/02.fp/data.017/aPOSCAR1/deepmd",
"./Y/relable/iter.000006/02.fp/data.017/aPOSCAR2/deepmd",
"./Y/relable/iter.000006/02.fp/data.017/aPOSCAR3/deepmd",
"./Y/relable/iter.000006/02.fp/data.017/aPOSCAR4/deepmd",
"./Y/relable/iter.000006/02.fp/data.017/aPOSCAR5/deepmd",
"./Y/relable/iter.000006/02.fp/data.018/aPOSCAR0/deepmd",
"./Y/relable/iter.000006/02.fp/data.018/aPOSCAR1/deepmd",
"./Y/relable/iter.000006/02.fp/data.018/aPOSCAR2/deepmd",
"./Y/relable/iter.000006/02.fp/data.018/aPOSCAR3/deepmd",
"./Y/relable/iter.000006/02.fp/data.018/aPOSCAR4/deepmd",
"./Y/relable/iter.000006/02.fp/data.018/aPOSCAR5/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR0/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR1/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR10/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR11/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR12/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR2/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR3/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR4/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR5/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR6/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR7/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR8/deepmd",
"./Y/relable/iter.000006/02.fp/data.019/aPOSCAR9/deepmd",
"./Y/relable/iter.000007/02.fp/data.004/aPOSCAR0/deepmd",
"./Y/relable/iter.000007/02.fp/data.006/aPOSCAR0/deepmd",
"./Y/relable/iter.000007/02.fp/data.006/aPOSCAR1/deepmd",
"./Y/relable/iter.000007/02.fp/data.007/aPOSCAR0/deepmd",
"./Y/relable/iter.000007/02.fp/data.007/aPOSCAR1/deepmd",
"./Y/relable/iter.000007/02.fp/data.013/aPOSCAR0/deepmd",
"./Y/relable/iter.000007/02.fp/data.013/aPOSCAR1/deepmd",
"./Y/relable/iter.000007/02.fp/data.020/aPOSCAR0/deepmd",
"./Y/relable/iter.000007/02.fp/data.020/aPOSCAR1/deepmd",
"./Y/relable/iter.000007/02.fp/data.022/aPOSCAR0/deepmd",
"./Y/relable/iter.000007/02.fp/data.022/aPOSCAR1/deepmd",
"./Y/relable/iter.000007/02.fp/data.023/aPOSCAR0/deepmd",
"./Y/relable/iter.000007/02.fp/data.023/aPOSCAR1/deepmd",
"./Y/relable/iter.000007/02.fp/data.023/aPOSCAR2/deepmd",
"./Y/relable/iter.000007/02.fp/data.023/aPOSCAR3/deepmd",
"./Y/relable/iter.000007/02.fp/data.023/aPOSCAR4/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR0/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR1/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR10/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR11/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR12/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR13/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR14/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR2/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR3/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR4/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR5/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR6/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR7/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR8/deepmd",
"./Y/relable/iter.000008/02.fp/data.000/aPOSCAR9/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR0/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR1/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR10/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR11/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR12/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR13/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR2/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR3/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR4/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR5/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR6/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR7/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR8/deepmd",
"./Y/relable/iter.000008/02.fp/data.008/aPOSCAR9/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR0/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR1/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR10/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR11/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR12/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR13/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR2/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR3/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR4/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR5/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR6/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR7/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR8/deepmd",
"./Y/relable/iter.000008/02.fp/data.016/aPOSCAR9/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR0/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR1/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR10/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR11/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR12/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR13/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR2/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR3/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR4/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR5/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR6/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR7/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR8/deepmd",
"./Y/relable/iter.000009/02.fp/data.001/aPOSCAR9/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR0/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR1/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR10/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR11/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR12/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR13/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR14/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR2/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR3/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR4/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR5/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR6/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR7/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR8/deepmd",
"./Y/relable/iter.000009/02.fp/data.009/aPOSCAR9/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR0/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR1/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR10/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR11/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR12/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR13/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR2/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR3/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR4/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR5/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR6/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR7/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR8/deepmd",
"./Y/relable/iter.000009/02.fp/data.017/aPOSCAR9/deepmd",

        "./Mg-Y/bcc/1/POSCAR.01x01x01/02.md/sys-0012-0004/deepmd",
	"./Mg-Y/bcc/2/POSCAR.01x01x01/02.md/sys-0013-0003/deepmd",
	"./Mg-Y/bcc/3/POSCAR.01x01x01/02.md/sys-0014-0002/deepmd",
	"./Mg-Y/bcc/4/POSCAR.01x01x01/02.md/sys-0015-0001/deepmd",
	"./Mg-Y/fcc/1/POSCAR.01x01x01/02.md/sys-0023-0009/deepmd",
	"./Mg-Y/fcc/2/POSCAR.01x01x01/02.md/sys-0028-0004/deepmd",
	"./Mg-Y/fcc/3/POSCAR.01x01x01/02.md/sys-0031-0001/deepmd",
	"./Mg-Y/hcp/1/POSCAR.01x01x01/02.md/sys-0015-0001/deepmd",
	"./Mg-Y/hcp/2/POSCAR.01x01x01/02.md/sys-0014-0002/deepmd",
	"./Mg-Y/hcp/3/POSCAR.01x01x01/02.md/sys-0013-0003/deepmd",
	"./Mg-Y/hcp/4/POSCAR.01x01x01/02.md/sys-0014-0002/deepmd",
	"./Mg-Y/hcp/5/POSCAR.01x01x01/02.md/sys-0014-0002/deepmd",
	"./Mg-Y/hcp/6/POSCAR.01x01x01/02.md/sys-0012-0004/deepmd",
	"./Mg-Y/compounds/1/POSCAR.02x02x01/02.md/sys-0016-0008/deepmd",
        "./Mg-Y/compounds/2/POSCAR.02x01x01/02.md/sys-0016-0008/deepmd",
	"./Mg-Y/compounds/3/POSCAR.02x02x02/02.md/sys-0008-0008/deepmd",
        "./Mg-Y/compounds/4/POSCAR.01x01x01/02.md/sys-0048-0010/deepmd"

    ],
    
    "sys_configs_prefix": "/data/wangyinan/wangyinan/Mg-Y/init",
    "sys_configs": [

	    ["Mg-Y/bcc/1/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-0.970/00000[0-4]/POSCAR"],
	    ["Mg-Y/bcc/1/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/bcc/1/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.030/00000[0-4]/POSCAR"],
	    ["Mg-Y/bcc/2/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.970/00000[0-4]/POSCAR"],
	    ["Mg-Y/bcc/2/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/bcc/2/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.030/00000[0-4]/POSCAR"],
	    ["Mg-Y/bcc/3/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.970/00000[0-4]/POSCAR"],
	    ["Mg-Y/bcc/3/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/bcc/3/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.030/00000[0-4]/POSCAR"],
	    ["Mg-Y/bcc/4/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.970/00000[0-4]/POSCAR"],
	    ["Mg-Y/bcc/4/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/bcc/4/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.030/00000[0-4]/POSCAR"],
	    ["Mg-Y/compounds/1/POSCAR.02x02x01/01.scale_pert/sys-0016-0008/scale-0.970/00000[0-4]/POSCAR"],
	    ["Mg-Y/compounds/1/POSCAR.02x02x01/01.scale_pert/sys-0016-0008/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/compounds/1/POSCAR.02x02x01/01.scale_pert/sys-0016-0008/scale-1.030/00000[0-4]/POSCAR"],
	    ["Mg-Y/compounds/2/POSCAR.02x01x01/01.scale_pert/sys-0016-0008/scale-0.970/00000[0-4]/POSCAR"],
	    ["Mg-Y/compounds/2/POSCAR.02x01x01/01.scale_pert/sys-0016-0008/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/compounds/2/POSCAR.02x01x01/01.scale_pert/sys-0016-0008/scale-1.030/00000[0-4]/POSCAR"],
	    ["Mg-Y/compounds/3/POSCAR.02x02x02/01.scale_pert/sys-0008-0008/scale-0.970/00000[0-4]/POSCAR"],
	    ["Mg-Y/compounds/3/POSCAR.02x02x02/01.scale_pert/sys-0008-0008/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/compounds/3/POSCAR.02x02x02/01.scale_pert/sys-0008-0008/scale-1.030/00000[0-4]/POSCAR"],
	    ["Mg-Y/compounds/4/POSCAR.01x01x01/01.scale_pert/sys-0048-0010/scale-0.970/00000[0-4]/POSCAR"],
	    ["Mg-Y/compounds/4/POSCAR.01x01x01/01.scale_pert/sys-0048-0010/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/compounds/4/POSCAR.01x01x01/01.scale_pert/sys-0048-0010/scale-1.030/00000[0-4]/POSCAR"],
	    ["Mg-Y/fcc/1/POSCAR.01x01x01/01.scale_pert/sys-0023-0009/scale-0.970/00000[0-4]/POSCAR"],
	    ["Mg-Y/fcc/1/POSCAR.01x01x01/01.scale_pert/sys-0023-0009/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/fcc/1/POSCAR.01x01x01/01.scale_pert/sys-0023-0009/scale-1.030/00000[0-4]/POSCAR"],
	    ["Mg-Y/fcc/2/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-0.970/00000[0-4]/POSCAR"],
	    ["Mg-Y/fcc/2/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/fcc/2/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.030/00000[0-4]/POSCAR"],
	    ["Mg-Y/fcc/3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-0.970/00000[0-4]/POSCAR"],
	    ["Mg-Y/fcc/3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/fcc/3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.030/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.960/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.980/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.020/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.040/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.960/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.980/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.020/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.040/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.960/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.980/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.020/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.040/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.960/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.980/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.020/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.040/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.960/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.980/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.020/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.040/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-0.960/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-0.980/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.000/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.020/00000[0-4]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.040/00000[0-4]/POSCAR"],

	    ["Mg-Y/bcc/1/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-0.970/00000[5-9]/POSCAR"],
	    ["Mg-Y/bcc/1/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/bcc/1/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.030/00000[5-9]/POSCAR"],
	    ["Mg-Y/bcc/2/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.970/00000[5-9]/POSCAR"],
	    ["Mg-Y/bcc/2/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/bcc/2/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.030/00000[5-9]/POSCAR"],
	    ["Mg-Y/bcc/3/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.970/00000[5-9]/POSCAR"],
	    ["Mg-Y/bcc/3/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/bcc/3/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.030/00000[5-9]/POSCAR"],
	    ["Mg-Y/bcc/4/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.970/00000[5-9]/POSCAR"],
	    ["Mg-Y/bcc/4/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/bcc/4/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.030/00000[5-9]/POSCAR"],
	    ["Mg-Y/compounds/1/POSCAR.02x02x01/01.scale_pert/sys-0016-0008/scale-0.970/00000[5-9]/POSCAR"],
	    ["Mg-Y/compounds/1/POSCAR.02x02x01/01.scale_pert/sys-0016-0008/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/compounds/1/POSCAR.02x02x01/01.scale_pert/sys-0016-0008/scale-1.030/00000[5-9]/POSCAR"],
	    ["Mg-Y/compounds/2/POSCAR.02x01x01/01.scale_pert/sys-0016-0008/scale-0.970/00000[5-9]/POSCAR"],
	    ["Mg-Y/compounds/2/POSCAR.02x01x01/01.scale_pert/sys-0016-0008/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/compounds/2/POSCAR.02x01x01/01.scale_pert/sys-0016-0008/scale-1.030/00000[5-9]/POSCAR"],
	    ["Mg-Y/compounds/3/POSCAR.02x02x02/01.scale_pert/sys-0008-0008/scale-0.970/00000[5-9]/POSCAR"],
	    ["Mg-Y/compounds/3/POSCAR.02x02x02/01.scale_pert/sys-0008-0008/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/compounds/3/POSCAR.02x02x02/01.scale_pert/sys-0008-0008/scale-1.030/00000[5-9]/POSCAR"],
	    ["Mg-Y/compounds/4/POSCAR.01x01x01/01.scale_pert/sys-0048-0010/scale-0.970/00000[5-9]/POSCAR"],
	    ["Mg-Y/compounds/4/POSCAR.01x01x01/01.scale_pert/sys-0048-0010/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/compounds/4/POSCAR.01x01x01/01.scale_pert/sys-0048-0010/scale-1.030/00000[5-9]/POSCAR"],
	    ["Mg-Y/fcc/1/POSCAR.01x01x01/01.scale_pert/sys-0023-0009/scale-0.970/00000[5-9]/POSCAR"],
	    ["Mg-Y/fcc/1/POSCAR.01x01x01/01.scale_pert/sys-0023-0009/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/fcc/1/POSCAR.01x01x01/01.scale_pert/sys-0023-0009/scale-1.030/00000[5-9]/POSCAR"],
	    ["Mg-Y/fcc/2/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-0.970/00000[5-9]/POSCAR"],
	    ["Mg-Y/fcc/2/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/fcc/2/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.030/00000[5-9]/POSCAR"],
	    ["Mg-Y/fcc/3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-0.970/00000[5-9]/POSCAR"],
	    ["Mg-Y/fcc/3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/fcc/3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.030/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.960/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.980/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.020/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.040/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.960/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.980/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.020/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.040/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.960/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.980/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.020/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.040/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.960/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.980/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.020/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.040/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.960/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.980/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.020/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.040/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-0.960/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-0.980/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.000/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.020/00000[5-9]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.040/00000[5-9]/POSCAR"],

	    ["Mg-Y/bcc/1/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-0.970/00001[0-4]/POSCAR"],
	    ["Mg-Y/bcc/1/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/bcc/1/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.030/00001[0-4]/POSCAR"],
	    ["Mg-Y/bcc/2/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.970/00001[0-4]/POSCAR"],
	    ["Mg-Y/bcc/2/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/bcc/2/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.030/00001[0-4]/POSCAR"],
	    ["Mg-Y/bcc/3/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.970/00001[0-4]/POSCAR"],
	    ["Mg-Y/bcc/3/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/bcc/3/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.030/00001[0-4]/POSCAR"],
	    ["Mg-Y/bcc/4/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.970/00001[0-4]/POSCAR"],
	    ["Mg-Y/bcc/4/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/bcc/4/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.030/00001[0-4]/POSCAR"],
	    ["Mg-Y/compounds/1/POSCAR.02x02x01/01.scale_pert/sys-0016-0008/scale-0.970/00001[0-4]/POSCAR"],
	    ["Mg-Y/compounds/1/POSCAR.02x02x01/01.scale_pert/sys-0016-0008/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/compounds/1/POSCAR.02x02x01/01.scale_pert/sys-0016-0008/scale-1.030/00001[0-4]/POSCAR"],
	    ["Mg-Y/compounds/2/POSCAR.02x01x01/01.scale_pert/sys-0016-0008/scale-0.970/00001[0-4]/POSCAR"],
	    ["Mg-Y/compounds/2/POSCAR.02x01x01/01.scale_pert/sys-0016-0008/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/compounds/2/POSCAR.02x01x01/01.scale_pert/sys-0016-0008/scale-1.030/00001[0-4]/POSCAR"],
	    ["Mg-Y/compounds/3/POSCAR.02x02x02/01.scale_pert/sys-0008-0008/scale-0.970/00001[0-4]/POSCAR"],
	    ["Mg-Y/compounds/3/POSCAR.02x02x02/01.scale_pert/sys-0008-0008/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/compounds/3/POSCAR.02x02x02/01.scale_pert/sys-0008-0008/scale-1.030/00001[0-4]/POSCAR"],
	    ["Mg-Y/compounds/4/POSCAR.01x01x01/01.scale_pert/sys-0048-0010/scale-0.970/00001[0-4]/POSCAR"],
	    ["Mg-Y/compounds/4/POSCAR.01x01x01/01.scale_pert/sys-0048-0010/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/compounds/4/POSCAR.01x01x01/01.scale_pert/sys-0048-0010/scale-1.030/00001[0-4]/POSCAR"],
	    ["Mg-Y/fcc/1/POSCAR.01x01x01/01.scale_pert/sys-0023-0009/scale-0.970/00001[0-4]/POSCAR"],
	    ["Mg-Y/fcc/1/POSCAR.01x01x01/01.scale_pert/sys-0023-0009/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/fcc/1/POSCAR.01x01x01/01.scale_pert/sys-0023-0009/scale-1.030/00001[0-4]/POSCAR"],
	    ["Mg-Y/fcc/2/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-0.970/00001[0-4]/POSCAR"],
	    ["Mg-Y/fcc/2/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/fcc/2/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.030/00001[0-4]/POSCAR"],
	    ["Mg-Y/fcc/3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-0.970/00001[0-4]/POSCAR"],
	    ["Mg-Y/fcc/3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/fcc/3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.030/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.960/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.980/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.020/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.040/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.960/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.980/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.020/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.040/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.960/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.980/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.020/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.040/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.960/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.980/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.020/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.040/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.960/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.980/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.020/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.040/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-0.960/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-0.980/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.000/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.020/00001[0-4]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.040/00001[0-4]/POSCAR"],

	    ["Mg-Y/bcc/1/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-0.970/00001[5-9]/POSCAR"],
	    ["Mg-Y/bcc/1/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/bcc/1/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.030/00001[5-9]/POSCAR"],
	    ["Mg-Y/bcc/2/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.970/00001[5-9]/POSCAR"],
	    ["Mg-Y/bcc/2/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/bcc/2/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.030/00001[5-9]/POSCAR"],
	    ["Mg-Y/bcc/3/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.970/00001[5-9]/POSCAR"],
	    ["Mg-Y/bcc/3/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/bcc/3/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.030/00001[5-9]/POSCAR"],
	    ["Mg-Y/bcc/4/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.970/00001[5-9]/POSCAR"],
	    ["Mg-Y/bcc/4/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/bcc/4/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.030/00001[5-9]/POSCAR"],
	    ["Mg-Y/compounds/1/POSCAR.02x02x01/01.scale_pert/sys-0016-0008/scale-0.970/00001[5-9]/POSCAR"],
	    ["Mg-Y/compounds/1/POSCAR.02x02x01/01.scale_pert/sys-0016-0008/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/compounds/1/POSCAR.02x02x01/01.scale_pert/sys-0016-0008/scale-1.030/00001[5-9]/POSCAR"],
	    ["Mg-Y/compounds/2/POSCAR.02x01x01/01.scale_pert/sys-0016-0008/scale-0.970/00001[5-9]/POSCAR"],
	    ["Mg-Y/compounds/2/POSCAR.02x01x01/01.scale_pert/sys-0016-0008/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/compounds/2/POSCAR.02x01x01/01.scale_pert/sys-0016-0008/scale-1.030/00001[5-9]/POSCAR"],
	    ["Mg-Y/compounds/3/POSCAR.02x02x02/01.scale_pert/sys-0008-0008/scale-0.970/00001[5-9]/POSCAR"],
	    ["Mg-Y/compounds/3/POSCAR.02x02x02/01.scale_pert/sys-0008-0008/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/compounds/3/POSCAR.02x02x02/01.scale_pert/sys-0008-0008/scale-1.030/00001[5-9]/POSCAR"],
	    ["Mg-Y/compounds/4/POSCAR.01x01x01/01.scale_pert/sys-0048-0010/scale-0.970/00001[5-9]/POSCAR"],
	    ["Mg-Y/compounds/4/POSCAR.01x01x01/01.scale_pert/sys-0048-0010/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/compounds/4/POSCAR.01x01x01/01.scale_pert/sys-0048-0010/scale-1.030/00001[5-9]/POSCAR"],
	    ["Mg-Y/fcc/1/POSCAR.01x01x01/01.scale_pert/sys-0023-0009/scale-0.970/00001[5-9]/POSCAR"],
	    ["Mg-Y/fcc/1/POSCAR.01x01x01/01.scale_pert/sys-0023-0009/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/fcc/1/POSCAR.01x01x01/01.scale_pert/sys-0023-0009/scale-1.030/00001[5-9]/POSCAR"],
	    ["Mg-Y/fcc/2/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-0.970/00001[5-9]/POSCAR"],
	    ["Mg-Y/fcc/2/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/fcc/2/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.030/00001[5-9]/POSCAR"],
	    ["Mg-Y/fcc/3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-0.970/00001[5-9]/POSCAR"],
	    ["Mg-Y/fcc/3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/fcc/3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.030/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.960/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.980/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.020/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/1/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.040/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.960/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.980/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.020/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.040/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.960/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.980/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.020/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.040/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.960/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.980/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.020/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/4/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.040/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.960/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.980/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.020/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/5/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.040/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-0.960/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-0.980/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.000/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.020/00001[5-9]/POSCAR"],
	    ["Mg-Y/hcp/6/POSCAR.01x01x01/01.scale_pert/sys-0012-0004/scale-1.040/00001[5-9]/POSCAR"],
	    ["surf/1-0/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/1-1/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/1-2/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/1-3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/2-0/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/2-1/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/2-2/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/2-3/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/3-0/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/3-1/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/3-2/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/3-3/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/4-0/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/4-1/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/4-2/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/4-3/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/5-0/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/5-1/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/5-2/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/5-3/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/0/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/0/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/0/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/1/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/1/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/1/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/4/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/4/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/4/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/5/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/5/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/5/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/6/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/6/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/6/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/7/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/7/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/7/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/8/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/8/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/8/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/9/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/9/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/9/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/10/POSCAR.01x01x01/01.scale_pert/sys-0023-0001/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/10/POSCAR.01x01x01/01.scale_pert/sys-0023-0001/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/10/POSCAR.01x01x01/01.scale_pert/sys-0023-0001/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/11/POSCAR.01x01x01/01.scale_pert/sys-0022-0002/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/11/POSCAR.01x01x01/01.scale_pert/sys-0022-0002/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/11/POSCAR.01x01x01/01.scale_pert/sys-0022-0002/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/12/POSCAR.01x01x01/01.scale_pert/sys-0022-0002/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/12/POSCAR.01x01x01/01.scale_pert/sys-0022-0002/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/12/POSCAR.01x01x01/01.scale_pert/sys-0022-0002/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/13/POSCAR.01x01x01/01.scale_pert/sys-0021-0003/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/13/POSCAR.01x01x01/01.scale_pert/sys-0021-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/13/POSCAR.01x01x01/01.scale_pert/sys-0021-0003/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/14/POSCAR.01x01x01/01.scale_pert/sys-0021-0003/scale-0.970/00000[0-9]/POSCAR"],
	    ["surf/14/POSCAR.01x01x01/01.scale_pert/sys-0021-0003/scale-1.000/00000[0-9]/POSCAR"],
	    ["surf/14/POSCAR.01x01x01/01.scale_pert/sys-0021-0003/scale-1.030/00000[0-9]/POSCAR"],
	    ["surf/1-0/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/1-1/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/1-2/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/1-3/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/2-0/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/2-1/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/2-2/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/2-3/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/3-0/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/3-1/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/3-2/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/3-3/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/4-0/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/4-1/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/4-2/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/4-3/POSCAR.01x01x01/01.scale_pert/sys-0030-0002/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/5-0/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/5-1/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/5-2/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/5-3/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/0/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/0/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/0/POSCAR.01x01x01/01.scale_pert/sys-0015-0001/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/1/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/1/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/1/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/2/POSCAR.01x01x01/01.scale_pert/sys-0014-0002/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/3/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/4/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/4/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/4/POSCAR.01x01x01/01.scale_pert/sys-0013-0003/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/5/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/5/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/5/POSCAR.01x01x01/01.scale_pert/sys-0031-0001/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/6/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/6/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/6/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/7/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/7/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/7/POSCAR.01x01x01/01.scale_pert/sys-0029-0003/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/8/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/8/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/8/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/9/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/9/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/9/POSCAR.01x01x01/01.scale_pert/sys-0028-0004/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/10/POSCAR.01x01x01/01.scale_pert/sys-0023-0001/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/10/POSCAR.01x01x01/01.scale_pert/sys-0023-0001/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/10/POSCAR.01x01x01/01.scale_pert/sys-0023-0001/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/11/POSCAR.01x01x01/01.scale_pert/sys-0022-0002/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/11/POSCAR.01x01x01/01.scale_pert/sys-0022-0002/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/11/POSCAR.01x01x01/01.scale_pert/sys-0022-0002/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/12/POSCAR.01x01x01/01.scale_pert/sys-0022-0002/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/12/POSCAR.01x01x01/01.scale_pert/sys-0022-0002/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/12/POSCAR.01x01x01/01.scale_pert/sys-0022-0002/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/13/POSCAR.01x01x01/01.scale_pert/sys-0021-0003/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/13/POSCAR.01x01x01/01.scale_pert/sys-0021-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/13/POSCAR.01x01x01/01.scale_pert/sys-0021-0003/scale-1.030/00001[0-9]/POSCAR"],
	    ["surf/14/POSCAR.01x01x01/01.scale_pert/sys-0021-0003/scale-0.970/00001[0-9]/POSCAR"],
	    ["surf/14/POSCAR.01x01x01/01.scale_pert/sys-0021-0003/scale-1.000/00001[0-9]/POSCAR"],
	    ["surf/14/POSCAR.01x01x01/01.scale_pert/sys-0021-0003/scale-1.030/00001[0-9]/POSCAR"]
    ],

    "_comment":         " 00.train ",
    "training_init_model": true,
    "training_reuse_iter":              36,
    "training_reuse_old_ratio":         0.9,
    "training_reuse_start_lr":          1e-4,
    "training_reuse_stop_batch":        600000,
    "training_reuse_start_pref_e":      1,
    "training_reuse_start_pref_f":      100,
    "training_reuse_start_pref_v":      0,

    "numb_models":      4,
    "train_param":      "input.json",

    "default_training_param" :{
     "model": {
      "descriptor" :{
      "type":            "se_a",
      "sel":             [500,500],
      "rcut_smth":       2.00,
      "rcut":            8.00,
      "neuron":          [25,50,100],
      "resnet_dt":       false,
      "axis_neuron":     12,
      "seed":            0
      },

     "fitting_net" :{
      "n_neuron":         [240,240,240],
      "resnet_dt":        true,
      "coord_norm":       true,
      "type_fitting_net": false,
      "seed":             0
      }
     },

    "learning_rate" :{
     "type":              "exp",
     "start_lr":          0.0001,
     "decay_steps":       3000,
     "decay_rate":        0.95
     },

    "loss" : {
     "start_pref_e":      1,
     "limit_pref_e":      2,
     "start_pref_f":      100,
     "limit_pref_f":      1,
     "start_pref_v":      0.0,
     "limit_pref_v":      0.0
     },

    "training" : {
     "systems":          [],
     "set_prefix":       "set",
     "stop_batch":       600000,
     "batch_size":       1,

     "seed":             0,

     "_comment": " display and restart",
     "_comment": " frequencies counted in batch",
     "disp_file":        "lcurve.out",
     "disp_freq":        2000,
     "numb_test":        2,
     "save_freq":        2000,
     "save_ckpt":        "model.ckpt",
     "load_ckpt":        "model.ckpt",
     "disp_training":    true,
     "time_training":    true,
     "profiling":        false,
     "profiling_file":   "timeline.json",
     "_comment":         "that's all"
     }
          },

    "_comment":         " 01.model_devi ",
    "_comment": "model_devi_skip: the first x of the recorded frames",
    "model_devi_dt":            0.002,
    "model_devi_skip":          0,
    "model_devi_f_trust_lo":    0.05,
    "model_devi_f_trust_hi":    0.20,
    "model_devi_e_trust_lo":    1e10,
    "model_devi_e_trust_hi":    1e10,
    "model_devi_clean_traj":    false,
    "model_devi_jobs":
[
        {
            "_idx": 0,
            "ensemble": "npt-tri",
            "nsteps": 500,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62 ],
            "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
            "trj_freq": 10
        },
        {
            "_idx": 1,
            "ensemble": "npt-tri",
            "nsteps": 500,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125  ],
            "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
            "trj_freq": 10
        },

        {
            "_idx": 2,
            "ensemble": "npt-tri",
            "nsteps": 1000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188  ],
            "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
            "trj_freq": 10
        },

        {
            "_idx": 3,
            "ensemble": "npt-tri",
            "nsteps": 1000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251  ],
            "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
            "trj_freq": 10
        },

        {
            "_idx": 4,
            "ensemble": "npt-tri",
            "nsteps": 3000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62 ],
            "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
            "trj_freq": 10
        },
        {
            "_idx": 5,
            "ensemble": "npt-tri",
            "nsteps": 3000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125  ],
            "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
            "trj_freq": 10
        },

        {
            "_idx": 6,
            "ensemble": "npt-tri",
            "nsteps": 6000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188  ],
            "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
            "trj_freq": 10
        },

        {
            "_idx": 7,
            "ensemble": "npt-tri",
            "nsteps": 6000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251  ],
            "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
            "trj_freq": 10
        },

        {
            "_idx": 8,
            "ensemble": "npt-tri",
            "nsteps": 500,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62 ],
            "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
            "trj_freq": 10
        },
        {
            "_idx": 9,
            "ensemble": "npt-tri",
            "nsteps": 500,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125  ],
            "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
            "trj_freq": 10
        },

        {
            "_idx": 10,
            "ensemble": "npt-tri",
            "nsteps": 1000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188  ],
            "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
            "trj_freq": 10
        },

        {
            "_idx": 11,
            "ensemble": "npt-tri",
            "nsteps": 1000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251  ],
            "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
            "trj_freq": 10
        },

        {
            "_idx": 12,
            "ensemble": "npt-tri",
            "nsteps": 3000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62 ],
            "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
            "trj_freq": 10
        },
        {
            "_idx": 13,
            "ensemble": "npt-tri",
            "nsteps": 3000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125  ],
            "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
            "trj_freq": 10
        },

        {
            "_idx": 14,
            "ensemble": "npt-tri",
            "nsteps": 6000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188  ],
            "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
            "trj_freq": 10
        },

        {
            "_idx": 15,
            "ensemble": "npt-tri",
            "nsteps": 6000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251  ],
            "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
            "trj_freq": 10
        },

        {
            "_idx": 16,
            "ensemble": "npt-tri",
            "nsteps": 500,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62 ],
            "temps": [ 1799, 1978.9, 2158.8, 2338.7, 2518.6 ],
            "trj_freq": 10
        },
        {
            "_idx": 17,
            "ensemble": "npt-tri",
            "nsteps": 500,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125  ],
            "temps": [ 1799, 1978.9, 2158.8, 2338.7, 2518.6 ],
            "trj_freq": 10
        },

        {
            "_idx": 18,
            "ensemble": "npt-tri",
            "nsteps": 1000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188  ],
            "temps": [ 1799, 1978.9, 2158.8, 2338.7, 2518.6 ],
            "trj_freq": 10
        },

        {
            "_idx": 19,
            "ensemble": "npt-tri",
            "nsteps": 1000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251  ],
            "temps": [ 1799, 1978.9, 2158.8, 2338.7, 2518.6 ],
            "trj_freq": 10
        },

        {
            "_idx": 20,
            "ensemble": "npt-tri",
            "nsteps": 3000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62 ],
            "temps": [ 1799, 1978.9, 2158.8, 2338.7, 2518.6 ],
            "trj_freq": 10
        },
        {
            "_idx": 21,
            "ensemble": "npt-tri",
            "nsteps": 3000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125  ],
            "temps": [ 1799, 1978.9, 2158.8, 2338.7, 2518.6 ],
            "trj_freq": 10
        },

        {
            "_idx": 22,
            "ensemble": "npt-tri",
            "nsteps": 3500,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188  ],
            "temps": [ 1799, 1978.9, 2158.8, 2338.7, 2518.6 ],
            "trj_freq": 10
        },

        {
            "_idx": 23,
            "ensemble": "npt",
            "nsteps": 3500,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251  ],
            "temps": [ 1799, 1978.9, 2158.8, 2338.7, 2518.6 ],
            "trj_freq": 10
        },

        {
            "_idx": 24,
            "ensemble": "npt",
            "nsteps": 500,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62 ],
            "temps": [ 2698.5, 2878.4, 3058.3, 3238.2, 3418.1 ],
            "trj_freq": 10
        },
        {
            "_idx": 25,
            "ensemble": "npt",
            "nsteps": 500,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125  ],
            "temps": [ 2698.5, 2878.4, 3058.3, 3238.2, 3418.1 ],
            "trj_freq": 10
        },

        {
            "_idx": 26,
            "ensemble": "npt",
            "nsteps": 1000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188  ],
            "temps": [ 2698.5, 2878.4, 3058.3, 3238.2, 3418.1 ],
            "trj_freq": 10
        },

        {
            "_idx": 27,
            "ensemble": "npt",
            "nsteps": 1000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251  ],
            "temps": [ 2698.5, 2878.4, 3058.3, 3238.2, 3418.1 ],
            "trj_freq": 10
        },

        {
            "_idx": 28,
            "ensemble": "npt",
            "nsteps": 3000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62 ],
            "temps": [ 2698.5, 2878.4, 3058.3, 3238.2, 3418.1 ],
            "trj_freq": 10
        },
        {
            "_idx": 29,
            "ensemble": "npt",
            "nsteps": 3000,
            "press": [ 1.0, 10.0, 100.0, 1000.0, 5000.0, 10000.0 ],
            "sys_idx": [ 63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125  ],
            "temps": [ 2698.5, 2878.4, 3058.3, 3238.2, 3418.1 ],
            "trj_freq": 10
        },
        {
	 "_idx": 30,
	 "ensemble": "nvt",
	 "nsteps": 500,
	 "press": [ -1 ],
	 "sys_idx": [252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280,281,282,283,284,285,286,287,288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,303,304,305,306,307,308,309,310,311,312,313,314,315,316 ],
	 "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
	 "trj_freq": 10
       	},

        {
         "_idx": 31,
         "ensemble": "nvt",
         "nsteps": 500,
         "press": [ -1 ],
         "sys_idx": [ 317,318,319,320,321,322,323,324,325,326,327,328,329,330,331,332,333,334,335,336,337,338,339,340,341,342,343,344,345,346,347,348,349,350,351,352,353,354,355,356,357,358,359,360,361,362,363,364,365,366,367,368,369,370,371,372,373,374,375,376,377,378,379,380,381 ],
         "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
         "trj_freq": 10
        },

        {
         "_idx": 32,
         "ensemble": "nvt",
         "nsteps": 1000,
         "press": [ -1 ],
         "sys_idx": [252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280,281,282,283,284,285,286,287,288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,303,304,305,306,307,308,309,310,311,312,313,314,315,316 ],
         "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
         "trj_freq": 10
        },

        {
         "_idx": 33,
         "ensemble": "nvt",
         "nsteps": 1000,
         "press": [ -1 ],
         "sys_idx": [ 317,318,319,320,321,322,323,324,325,326,327,328,329,330,331,332,333,334,335,336,337,338,339,340,341,342,343,344,345,346,347,348,349,350,351,352,353,354,355,356,357,358,359,360,361,362,363,364,365,366,367,368,369,370,371,372,373,374,375,376,377,378,379,380,381 ],
         "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
         "trj_freq": 10
        },

        {
         "_idx": 34,
         "ensemble": "nvt",
         "nsteps": 3000,
         "press": [ -1 ],
         "sys_idx": [252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280,281,282,283,284,285,286,287,288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,303,304,305,306,307,308,309,310,311,312,313,314,315,316 ],
         "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
         "trj_freq": 10
        },

        {
         "_idx": 35,
         "ensemble": "nvt",
         "nsteps": 3000,
         "press": [ -1 ],
         "sys_idx": [ 317,318,319,320,321,322,323,324,325,326,327,328,329,330,331,332,333,334,335,336,337,338,339,340,341,342,343,344,345,346,347,348,349,350,351,352,353,354,355,356,357,358,359,360,361,362,363,364,365,366,367,368,369,370,371,372,373,374,375,376,377,378,379,380,381 ],
         "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
         "trj_freq": 10
        },

        {
         "_idx": 36,
         "ensemble": "nvt",
         "nsteps": 6000,
         "press": [ -1 ],
         "sys_idx": [252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280,281,282,283,284,285,286,287,288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,303,304,305,306,307,308,309,310,311,312,313,314,315,316 ],
         "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
         "trj_freq": 10
        },

        {
         "_idx": 37,
         "ensemble": "nvt",
         "nsteps": 6000,
         "press": [ -1 ],
         "sys_idx": [ 317,318,319,320,321,322,323,324,325,326,327,328,329,330,331,332,333,334,335,336,337,338,339,340,341,342,343,344,345,346,347,348,349,350,351,352,353,354,355,356,357,358,359,360,361,362,363,364,365,366,367,368,369,370,371,372,373,374,375,376,377,378,379,380,381 ],
         "temps": [ 50, 179.9, 359.8, 539.7, 719.6 ],
         "trj_freq": 10
        },
        {
	 "_idx": 38,
	 "ensemble": "nvt",
	 "nsteps": 500,
	 "press": [ -1 ],
	 "sys_idx": [252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280,281,282,283,284,285,286,287,288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,303,304,305,306,307,308,309,310,311,312,313,314,315,316 ],
	 "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
	 "trj_freq": 10
       	},

        {
         "_idx": 39,
         "ensemble": "nvt",
         "nsteps": 500,
         "press": [ -1 ],
         "sys_idx": [ 317,318,319,320,321,322,323,324,325,326,327,328,329,330,331,332,333,334,335,336,337,338,339,340,341,342,343,344,345,346,347,348,349,350,351,352,353,354,355,356,357,358,359,360,361,362,363,364,365,366,367,368,369,370,371,372,373,374,375,376,377,378,379,380,381 ],
         "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
         "trj_freq": 10
        },

        {
         "_idx": 40,
         "ensemble": "nvt",
         "nsteps": 1000,
         "press": [ -1 ],
         "sys_idx": [252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280,281,282,283,284,285,286,287,288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,303,304,305,306,307,308,309,310,311,312,313,314,315,316 ],
         "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
         "trj_freq": 10
        },

        {
         "_idx": 41,
         "ensemble": "nvt",
         "nsteps": 1000,
         "press": [ -1 ],
         "sys_idx": [ 317,318,319,320,321,322,323,324,325,326,327,328,329,330,331,332,333,334,335,336,337,338,339,340,341,342,343,344,345,346,347,348,349,350,351,352,353,354,355,356,357,358,359,360,361,362,363,364,365,366,367,368,369,370,371,372,373,374,375,376,377,378,379,380,381 ],
         "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
         "trj_freq": 10
        },

        {
         "_idx": 42,
         "ensemble": "nvt",
         "nsteps": 3000,
         "press": [ -1 ],
         "sys_idx": [252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280,281,282,283,284,285,286,287,288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,303,304,305,306,307,308,309,310,311,312,313,314,315,316 ],
         "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
         "trj_freq": 10
        },

        {
         "_idx": 43,
         "ensemble": "nvt",
         "nsteps": 3000,
         "press": [ -1 ],
         "sys_idx": [ 317,318,319,320,321,322,323,324,325,326,327,328,329,330,331,332,333,334,335,336,337,338,339,340,341,342,343,344,345,346,347,348,349,350,351,352,353,354,355,356,357,358,359,360,361,362,363,364,365,366,367,368,369,370,371,372,373,374,375,376,377,378,379,380,381 ],
         "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
         "trj_freq": 10
        },

        {
         "_idx": 44,
         "ensemble": "nvt",
         "nsteps": 6000,
         "press": [ -1 ],
         "sys_idx": [252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280,281,282,283,284,285,286,287,288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,303,304,305,306,307,308,309,310,311,312,313,314,315,316 ],
         "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
         "trj_freq": 10
        },

        {
         "_idx": 45,
         "ensemble": "nvt",
         "nsteps": 6000,
         "press": [ -1 ],
         "sys_idx": [ 317,318,319,320,321,322,323,324,325,326,327,328,329,330,331,332,333,334,335,336,337,338,339,340,341,342,343,344,345,346,347,348,349,350,351,352,353,354,355,356,357,358,359,360,361,362,363,364,365,366,367,368,369,370,371,372,373,374,375,376,377,378,379,380,381 ],
         "temps": [ 899.5, 1079.4, 1259.3, 1439.2, 1619.1 ],
         "trj_freq": 10
        }


    ],
    "_commnet": "02.fp",
    "cvasp": false,
    "fp_style": "vasp",
    "shuffle_poscar": false,
    "fp_task_max": 15,
    "fp_task_min": 2,
    "failed_ratio": 0.1,
    "fp_accurate_threshold": 0.9995,
    "fp_pp_path": "/data/wangyinan/wangyinan/Mg-Y/files/",
    "fp_pp_files": [
        "POTCAR_Mg",
        "POTCAR_Y"
    ],
    "fp_incar": "./INCAR_scf",
    "fp_params": {},
    "_comment":         " that's all "
}

{
  "train": [
    {
      "machine": {
        "batch": "shell",
        "hostname": "",
        "password": "",
        "port": 22,
        "username": "root",
        "work_path": "/root/dpgen_work",
        "cloud_resources": {
	        "cloud_platform": "ali",
          "AccessKey_ID":"",
          "AccessKey_Secret":"",
          "regionID": "cn-shanghai",
          "zone":     ["d"],
          "img_name": "kit-1.2.0",          
          "address":  "public",
          "machine_type_price": [
            {"machine_type": "ecs.gn5-c4g1.xlarge", "price_limit": 4.0, "numb": 1, "priority": 0}
          ],
          "instance_name": "MgY_train_yinan", 
          "pay_strategy": "spot"
        }
      },
      "resources": {
        "numb_gpu": 1,              
        "numb_node": 1,
        "task_per_node": 4,
        "partition": "gpu",         
        "exclude_list": [],
        "mem_limit": 28,
        "source_list": [],
        "module_list": [],
        "time_limit": "23:0:0"
      },
      "command": "/root/deepmd-kit/bin/dp",
      "group_size": 1
    }
  ],

  "model_devi": [
    {
      "machine": {
        "batch": "shell",
        "hostname": "",
        "password": "",
        "port": 22,
        "username": "root",
        "work_path": "/root/dpgen_work",
        "cloud_resources": {
	        "cloud_platform": "ali",
          "AccessKey_ID":"",
          "AccessKey_Secret":"",
          "regionID": "cn-shanghai",
          "zone":   ["e"],
          "img_name": "kit-1.2.0",
          "address":  "public",
          "machine_type_price": [
            {"machine_type": "ecs.gn5-c4g1.xlarge", "price_limit": 4.0, "numb": 1, "priority": 0}
          ],
          "instance_name": "MgY_modevi_yinan",
          "pay_strategy": "spot"
        }
      },
      "resources": {
        "numb_gpu": 1,
        "task_per_node": 4,
        "partition": "gpu",
        "exclude_list": [],
        "mem_limit": 90,
        "source_list": [],
        "module_list": [],
        "time_limit": "23:0:0"
      },
      "command": "/root/deepmd-kit/bin/lmp",
      "group_size": 500
    }
  ],

  "fp": [
    {
      "machine": {
        "batch": "shell",
        "hostname": "",
        "password": "",
        "port": 22,
        "username": "root",
        "work_path": "/root/dpgen_work",
        "cloud_resources": {
	        "cloud_platform": "ali",
          "AccessKey_ID":"",
          "AccessKey_Secret":"",
          "regionID": "cn-beijing",
	  "zone":   ["e", "f", "g"],
          "address":  "public",
          "img_name": "vasp",
          "machine_type_price": [
            {"machine_type": "ecs.g5.8xlarge", "price_limit": 0.05, "numb": 32, "priority": 1}
          ],
          "instance_name": "MgY_fp_yinan", 
          "pay_strategy": "spot"
        }
      },
      "resources": {
        "allow_failure": true,
        "ratio_failue": 0.05,
        "task_per_node": 32,
        "with_mpi": true,
        "source_list": ["/opt/intel/parallel_studio_xe_2018/psxevars.sh"],
        "envs" : {"PATH" : "/root/deepmd-pkg/vasp.5.4.4/bin:$PATH"}
      },
        "command": "vasp_std",
        "group_size": 6
    }
  ]
}

Transfer-learning

What is transfer-learning

Transfer Learning is the reuse of a pre-trained model on a new problem. It’s currently very popular in deep learning because it can train deep neural networks with comparatively little data. In our case, transfer learning means re-training a pre-trained DP model to another one by a small dataset, where the size of the new dataset might be orders of magnitude smaller in comparison to that of the original dataset used to train the original model.

When to use transfer-learning

Transfer learning has several benefits, where the main advantages are saving training time, better performance of neural networks (in most cases), and not needing a lot of data. The central idea under transfer-learning is that the knowledge of a model learned from a large dataset can be reused. In DP models, the typical knowledge that can be reused is the embedding net, which encode local atomic configurations into descriptors. For example, the following are some possible scenarios where transfer-learning may be useful:

Initially, we may generate DFT dataset without van der Waals correction. However, after sometime, we may find that van der Waals correction is important and should be considered.
Initially, we may generate DFT dataset PBE functional. However, after sometime, we may find that the PBE functional is not accurate enough. Alternatively, we may find that the original kspacing or energy cutoff cannot meet the convergence criterion for some highly distorted configurations.
We have a big DFT dataset of a material and DP models trained on the dataset in hand, and need a model of another material that is similar to the material. For example, we have a dataset of HfO2 in the DP library, and need a model of ZrO2.
We have a big DFT dataset of a material and DP models trained on the dataset in hand, and need to train a model on other properties by using the DeePMD-kit software. Here, the property to be trained is assumed also depending on local atomic configurations.

How to implement transfer-learning

This tutorial will introduce how to implement potential energy surface (PES) transfer-learning by using the DP-GEN software. In DP-GEN (version > 0.8.0), the “simplify” module is designed for this purpose. Suppose that we have completed a typical DP-GEN flow, and obtained the DFT dataset and four DP models. The workflow of “simplify” is similar to a typical DP-GEN process: iteratively training the DP models with the (re-) labeled data (00.train), picking data according to prediction deviations between different models (01.model_devi), and (re-) labeling the picked data (02.fp). Repeat the iterations until convergence is achieved. Then, the relabeled new dataset that is sufficient to refine the DP model is successfully collected.

In the “simplify” mode, the first iteration can be viewed as the initialization process in the conventional DP-GEN process, where the 00. train and 01. model_devi are skipped, and some data are randomly picked in 02.fp to be relabeled. The goal of relabeling may be using a different functional, using a different pseudopotential, using different parameters to achieve higher precision, etc. From the second iteration on:

In the training step (00. train), the DP models are modified based on the relabeled data with “init-model” mode by freezing some parameters of the DP models. For example, the parameters of the whole embedding net and the hidden layers of the fitting net can be fixed, and only the parameters of the output layer of the fitting net are trainable. The trainable parameters can be set according to your demand.
In the exploration step (01. model_devi), the deviations between predictions by the modified models on the original dataset are evaluated. Some of the data points (e.g. at most 100) with model deviation exceeding a criterion are randomly selected for relabeling.
In the labeling step (02.fp), the selected data points are relabeled, and fed to the new dataset.

The iterations will stop unit no data is picked up.

Example: Ag-Au

In this example, we generate the Ag-Au alloy DP model based on DFT-PBE without van der Waals correction. In fact, the van der Waals dispersion correction is necessary to correctly predict lattice parameter and surface formation energy in Ag-Au alloy system. Besides, DFT-PBE-D3 gives qualitatively reasonable results for the surface adsorption behaviors, which is widely used in investigating Ag segregation at the Ag-Au surface, and adatom adsorption/diffusion on Ag or Au surfaces. Therefore, a DP model honest to DFT-PBE-D3 is required for predicting surface behavior at larger scales.

The original DFT-PBE dataset is collected by using command “dpgen collect”, making the dataset convenient to read by simplify parameter files. Here, the DFT-PBE dataset is generated from the previous DPGEN iteration. Then, start the process by using command

dpgen simplify simplify.json machine.json

where the parameter file “simplify.json” is similar to that of a typical DP-GEN process. The “simplify.json” file is shown and explained in the following. The process converged in only four iterations, where merely 144, 155 and 275 structures were selected to be relabeled for pure Ag, pure Au, and Ag-Au alloys, respectively. The transferred DP-PBE-D3 model agrees with DFT-PBE-D3.

By the way, it is really cheap to do the D3 correction, which only requires the atom coordinates and the box size as the input information. Some codes could do this:

https://chemie.uni-bonn.de/pctc/mulliken-center/software/dft-d3/get-the-current-version-of-dft-d3

https://github.com/MMunibas/PhysNet/blob/master/neural_network/grimme_d3/grimme_d3.py

https://github.com/loriab/dftd3

https://github.com/dftbplus/dftd3-lib

According to the scripts on the last website, we relabeled all the data with DFT-PBE-D3. Therefore, this example can also be done without transfer-learning. However, in other scenarios listed in the part of “when to use transfer learning”, transfer-learning is beneficial.

Examplary “simplify.json” of Ag-Au system:

{
    "type_map":    [ "Ag", "Au" ],
    "mass_map":    [ 108, 196.967 ],
    "init_data_prefix": "",
    "init_data_sys":  [],
    "pick_data":  "/Ag-Au/simp/collect",
# Use 'dp collect' to put all the data in one fold. 
    "sys_batch_size":  "auto",
    "training_iter0_model_path":  "/Ag-Au/final/00.train/00[0-3]",
# Locate the original DP model.
    "training_init_model":    true,
# Transfer learning is doing 'init-model' based on the original one. 
    "numb_models":  4,
    "train_param":  "input.json",
    "default_training_param" : {
        "_comment": " model parameters",
    "model": {
            "type_map":     ["Ag", "Au"],
            "descriptor" :{
        "type":             "se_a",
        "sel":              [150,150],
        "rcut_smth":        2.00,
        "rcut":             6.00,
        "neuron":           [25, 50, 100],
        "trainable":      false,
# Parameters of embedding net are not trained.
    "resnet_dt":        false,
    "axis_neuron":      12,
    "seed":             0,
                "_comment":         " that's all"
            },
            "fitting_net" : {
    "neuron":           [240, 240, 240],
    "resnet_dt":        true,
    "trainable":      [false, false, false, true],
# The parameters of the output layer of the fitting net are trainable. 
    "seed":             1,
    "_comment":         " that's all"
            },
            "_comment":     " that's all"
  },
  "learning_rate" :{
            "type":         "exp",
            "start_lr":     0.001,
            "stop_lr":      3.51e-8,
            "decay_steps":  2000,
            "_comment":     "that's all"
  },
  "loss" :{
            "start_pref_e": 100,
            "limit_pref_e": 100,
            "start_pref_f": 1,
            "limit_pref_f": 1,
            "start_pref_v": 0.9,
            "limit_pref_v": 1.0,
            "_comment":     " that's all"
  },
  "_comment": " traing controls",
  "training" : {
            "systems":      [],
            "set_prefix":   "set",
            "stop_batch":   400000,
            "batch_size":   2,
            "seed":         1,
            "_comment": " display and restart",
            "_comment": " frequencies counted in batch",
            "disp_file":    "lcurve.out",
            "disp_freq":    2000,
            "numb_test":    2,
            "save_freq":    2000,
            "save_ckpt":    "model.ckpt",
            "load_ckpt":    "model.ckpt",
            "disp_training":true,
            "time_training":true,
            "profiling":    false,
            "profiling_file":"timeline.json",
            "_comment":     "that's all"
},
  "_comment":         "that's all"
},
"fp_style": "vasp",
"fp_skip_bad_box":  "length_ratio:5;height_ratio:5",
"shuffle_poscar":  false,
"fp_task_max":  100,
# At most 100 data points are randomly picked and sent to the labeling step.
"fp_task_min" :  0,
# All the data picked in the first iteration will be sent to relabel. 
    "fp_pp_path":  "vasp_input",
    "fp_pp_files":  ["POTCAR_Ag", "POTCAR_Au"],
    "fp_incar":         "vasp_input/INCAR",

    "use_clusters": false,
    "labeled": false,
    "init_pick_number":100,
    "iter_pick_number":100,
    "e_trust_lo":1e10,
    "e_trust_hi":1e10,
    "f_trust_lo":0.20,
    "f_trust_hi":100.00,
# Once the mode_deviation of the data is higher than f_trust_lo, it is selected as a candidate. Since all the data is selected from the original data set, which is proper for constructing a good DP model,  we do not especially need to exclude the data with bad box or atom overlap.    
    "_comment": " that's all "
}

References

YiNan Wang et al 2022 Modelling Simul. Mater. Sci. Eng. 30 025003

Grimme S, Antony J, Ehrlich S and Krieg H 2010 J. Chem. Phys. 132 154104

Learning Resources

Here is the learning Resources:

Video Tutorials（Chinese)

Some Video Resources：

Basic theoretical courses:

DeePMD-kit and DP-GEN

Writing Tips

Hello volunteers, this docs tells you how to write articles for DeepModeling tutorials.

You can just follow 2 steps:

Write in markdown format and put it into proper directories.
Change index.rst to show your doc.

Write in Markdown and Put into proper directories.

You should learn how to write a markdown document. It is quite easy!

Here we recommend you 2 website:

You should know the proper directories.

Our Github Address is: https://github.com/deepmodeling/tutorials

All doc is in: “source” directories. According to your doc’s purpose, your doc can be put into 4 directories in “source”:

Tutorials: Telling beginners how to run Deepmodeling projects.
Casestudies: Some case telling people how to use Deepmodeling projects.
Resources: Other resources for learning.
QA: Some questions and answers.

After that, you should find the proper directories and put your docs.

For example, if you write a “methane.md” for case study, you can put it into “/source/CaseStudies/Gas-phase”.

Change indexs.rst to show your doc.

Then you should change the index.rst to show your doc.

You can learn rst format here: reStructuredText

In short, you can simply change index.rst in your parent directories.

For example, if you put “methane.md” into “/source/CaseStudies/Gas-phase”, you can find and change “index.rst” in “source/CaseStudies/Gas-phase”. All you should do is imitating such file. I believe you can do it!

If you want to learn more detailed information about how to build this website, you can check this:

Q & A

here is Q & A

Discussions and Feedbacks:

The tutorials need feedbacks from you. If you think some tutorials are confused, please write your feedbacks on our discussion board.

Working Now:

At present, we are writing tutorials for:

DeePMD-kit
DP-GEN

Another team focus on writing some brief materials about AI + Science for beginners, such as:

What is machine learning
What is MD(Molecular Dynamics)
The core concept about AI + Science

Welcome to DeepModeling Tutorial!

Using DeePMD-Kit

Introduction

Installation

Install off-line packages

Install with conda

Install with docker

Verify the installation

Theory

How to Setup a DeePMD-kit Training within 5 Minutes

Tutorial Data

Preparing Data

Training

Freeze/Compress the Model

Handson-Tutorial(v2.0.3)

Workflow of the DeePMD-kit

Example: a gas-phase methane molecule

Data preparation

input script

Train process

Model application

Summary

Using DP-GEN

Install DP-GEN

Install with pip

Install with source code

Verify the installation

Hands-on tutorial for DP-GEN (v0.10.6)

Workflow of the DP-GEN

Example: a gas phase methane molecule

Init

Run

Input files

param.json

Start run process

Results analysis

Simplify

Auto-test

structure relaxation

step1-make

step2-run

step3-post

property calculation

step1-make

step2-run

step3-post

Summary

Practical-Guidelines-for-DP

Practical Guidelines for DP

Know the Toolbox Well

Deep Potential

Configurational Space

Sampling Methods

Limitations and Risks

DP-GEN

DP Library

Know the Physical Nature of a System

Know the Boundaries of a Problem

Convergence-Test

Convergence Test

Gas-Phase

Simulation of the oxidation of methane

Step 1: Preparing the reference dataset

Step 2. Training the Deep Potential (DP)

Step 3: Freeze the model

Step 4: Running MD simulation based on the DP

Step 5: Analysis of the trajectory

Acknowledge

References

Mg-Y_alloy

Mg-Y alloy system

Introduction

Folder Structure

Preparing Files

Transfer-learning

Transfer-learning

What is transfer-learning

When to use transfer-learning

How to implement transfer-learning

Example: Ag-Au

step1-`make`

step2-`run`

step3-`post`

step1-`make`

step2-`run`

step3-`post`