Merge pull request #75 from JisuJung928/master

docs fix complete

README.md

@@ -6,7 +6,7 @@ If you use SIMPLE-NN, please cite this article:
 K. Lee, D. Yoo, W. Jeong, S. Han, SIMPLE-NN: An efficient package for training and executing neural-network interatomic potentials, *Computer Physics Communications* (2019), https://doi.org/10.1016/j.cpc.2019.04.014.
 
 Here do we describe minimal instruction to run the example of SIMPLE-NN
-If you want more information such as tuning parameters, please visit our online manual(https://mtcg.snu.ac.kr/doc/index.html)
+If you want more information such as tuning parameters, please visit our online manual(https://simple-nn.readthedocs.io)
 
 ## Installation
 SIMPLE-NN use Tensorflow and mpi4py(optional).

docs/examples/examples.rst

@@ -22,9 +22,9 @@ Generate NNP
 
 To generate NNP using symmetry function and neural network, 
 you need three types of input file (input.yaml, str_list, params_XX) 
-as described in :doc:`/tutorials/tutorial:md-simulation-with-lammps` section.
+as described in :doc:`/tutorials/tutorial` section.
 The example files except params_Si and params_O are introduced below.
-Detail of params_Si and params_O can be found in :doc:`/features/symmetry_function/symmetry_function:inputs` section.
+Detail of params_Si and params_O can be found in :doc:`/features/symmetry_function/symmetry_function` section.
 Input files introduced in this section can be found in 
 :code:`SIMPLE-NN/examples/SiO2/generate_NNP`.
 

docs/features/symmetry_function/symmetry_function.rst

@@ -101,7 +101,7 @@ Each parameter indicates (SF means symmetry function) ::
 
 The number inside the indicates the number of parameters.
 
-First column indicates the type of symmetry function. Currently, G2 (2), G4 (4), and G5 (5) are available (defined in the reference[#f1]_).
+First column indicates the type of symmetry function. Currently, G2 (2), G4 (4), and G5 (5) are available.
 
 Second and third column indicates the type index of neighbor atoms which starts from 1.
 (The order of type index need to be the same as the order of the :code:`atom_types` tag indicated in :code:`input.yaml`) 

docs/models/hdnn/hdnn.rst

@@ -53,13 +53,14 @@ Network related parameter
   If :code:`true`, both energy and force are used for training.
 
 * :code:`force_coeff`\: (float, default: 0.1)
-  Ratio of force_loss to energy_loss in total_loss
+  Ratio of force loss to energy loss in total loss
 
 * :code:`use_stress`\: (boolean, default: false)
   If :code:`true`, both energy and stress are used for training.
+  (Caution : The unit of stress RMSE written in LOG file is kbar.)
 
 * :code:`stress_coeff`\: (float, default: 0.00001)
-  Ratio of stress_loss to energy_loss in total_loss
+  Ratio of stress loss to energy loss in total loss
 
 * :code:`stddev`\: (float, default: 0.3)
   Standard deviation for weights initialization.

simple_nn/models/neural_network.py

@@ -42,7 +42,7 @@ def __init__(self):
                                           'params': dict(),
                                       },
                                       'use_force': True,
-                                      'use_stress': True,
+                                      'use_stress': False,
                                       'double_precision': True,
                                       'weight_initializer': {
                                           'type': 'truncated normal',
@@ -176,7 +176,6 @@ def _make_model(self):
         if self.inputs['continue'] == 'weights':
             saved_weights = read_lammps_potential('potential_saved')
 
-        #acti_func = 'selu'
         #acti_func = 'elu'
         #acti_func = 'sigmoid'
         #acti_func = 'tanh'