In this section, we will take $deepmd_source_dir/examples/water/se_e2_a/input.json as an example of the input file. ## 4.2.1. Learning rate The learning_rate section in input.json is given as follows  "learning_rate" :{ "type": "exp", "start_lr": 0.001, "stop_lr": 3.51e-8, "decay_steps": 5000, "_comment": "that's all" }  • start_lr gives the learning rate at the beginning of the training. • stop_lr gives the learning rate at the end of the training. It should be small enough to ensure that the network parameters satisfactorily converge. • During the training, the learning rate decays exponentially from start_lr to stop_lr following the formula: $\alpha(t) = \alpha_0 \lambda ^ { t / \tau }$ where $$t$$ is the training step, $$\alpha$$ is the learning rate, $$\alpha_0$$ is the starting learning rate (set by start_lr), $$\lambda$$ is the decay rate, and $$\tau$$ is the decay steps, i.e.  lr(t) = start_lr * decay_rate ^ ( t / decay_steps )   ## 4.2.2. Training parameters Other training parameters are given in the training section.  "training": { "training_data": { "systems": ["../data_water/data_0/", "../data_water/data_1/", "../data_water/data_2/"], "batch_size": "auto" }, "validation_data":{ "systems": ["../data_water/data_3"], "batch_size": 1, "numb_btch": 3 }, "mixed_precision": { "output_prec": "float32", "compute_prec": "float16" }, "numb_steps": 1000000, "seed": 1, "disp_file": "lcurve.out", "disp_freq": 100, "save_freq": 1000 }  The sections training_data and validation_data give the training dataset and validation dataset, respectively. Taking the training dataset for example, the keys are explained below: • systems provide paths of the training data systems. DeePMD-kit allows you to provide multiple systems with different numbers of atoms. This key can be a list or a str. • list: systems gives the training data systems. • str: systems should be a valid path. DeePMD-kit will recursively search all data systems in this path. • At each training step, DeePMD-kit randomly picks batch_size frame(s) from one of the systems. The probability of using a system is by default in proportion to the number of batches in the system. More options are available for automatically determining the probability of using systems. One can set the key auto_prob to • "prob_uniform" all systems are used with the same probability. • "prob_sys_size" the probability of using a system is proportional to its size (number of frames). • "prob_sys_size; sidx_0:eidx_0:w_0; sidx_1:eidx_1:w_1;..." the list of systems is divided into blocks. Block i has systems ranging from sidx_i to eidx_i. The probability of using a system from block i is proportional to w_i. Within one block, the probability of using a system is proportional to its size. • An example of using "auto_prob" is given below. The probability of using systems is 0.4, and the sum of the probabilities of using systems and systems is 0.6. If the number of frames in systems is twice of system, then the probability of using system is 0.4 and that of system is 0.2.  "training_data": { "systems": ["../data_water/data_0/", "../data_water/data_1/", "../data_water/data_2/"], "auto_prob": "prob_sys_size; 0:2:0.6; 2:3:0.4", "batch_size": "auto" }  • The probability of using systems can also be specified explicitly with key sys_probs which is a list having the length of the number of systems. For example  "training_data": { "systems": ["../data_water/data_0/", "../data_water/data_1/", "../data_water/data_2/"], "sys_probs": [0.5, 0.3, 0.2], "batch_size": "auto:32" }  • The key batch_size specifies the number of frames used to train or validate the model in a training step. It can be set to • list: the length of which is the same as the systems. The batch size of each system is given by the elements of the list. • int: all systems use the same batch size. • "auto": the same as "auto:32", see "auto:N" • "auto:N": automatically determines the batch size so that the batch_size times the number of atoms in the system is no less than N. • The key numb_batch in validate_data gives the number of batches of model validation. Note that the batches may not be from the same system The section mixed_precision specifies the mixed precision settings, which will enable the mixed precision training workflow for DeePMD-kit. The keys are explained below: • output_prec precision used in the output tensors, only float32 is supported currently. • compute_prec precision used in the computing tensors, only float16 is supported currently. Note there are several limitations about mixed precision training: • Only se_e2_a type descriptor is supported by the mixed precision training workflow. • The precision of the embedding net and the fitting net are forced to be set to float32. Other keys in the training section are explained below: • numb_steps The number of training steps. • seed The random seed for getting frames from the training data set. • disp_file The file for printing learning curve. • disp_freq The frequency of printing learning curve. Set in the unit of training steps • save_freq The frequency of saving checkpoint. ## 4.2.3. Options and environment variables Several command line options can be passed to dp train, which can be checked with $ dp train --help


An explanation will be provided

positional arguments:
INPUT                 the input json database

optional arguments:
-h, --help            show this help message and exit

--init-model INIT_MODEL
Initialize a model by the provided checkpoint

--restart RESTART     Restart the training from the provided checkpoint

--init-frz-model INIT_FRZ_MODEL
Initialize the training from the frozen model.
--skip-neighbor-stat  Skip calculating neighbor statistics. Sel checking, automatic sel, and model compression will be disabled. (default: False)


--init-model model.ckpt, initializes the model training with an existing model that is stored in the path prefix of checkpoint files model.ckpt, the network architectures should match.

--restart model.ckpt, continues the training from the checkpoint model.ckpt.

--init-frz-model frozen_model.pb, initializes the training with an existing model that is stored in frozen_model.pb.

--skip-neighbor-stat will skip calculating neighbor statistics if one is concerned about performance. Some features will be disabled.

To maximize the performance, one should follow FAQ: How to control the parallelism of a job to control the number of threads.

One can set other environmental variables:

Environment variables

Allowed value

Default value

Usage

DP_INTERFACE_PREC

high, low

high

Control high (double) or low (float) precision of training.

DP_AUTO_PARALLELIZATION

0, 1

0

Enable auto parallelization for CPU operators.

DP_JIT

0, 1

0

Enable JIT. Note that this option may either improve or decrease the performance. Requires TensorFlow supports JIT.

## 4.2.4. Adjust sel of a frozen model

One can use --init-frz-model features to adjust (increase or decrease) sel of a existing model. Firstly, one needs to adjust sel in input.json. For example, adjust from [46, 92] to [23, 46].

"model": {
"descriptor": {
"sel": [23, 46]
}
}


To obtain the new model at once, numb_steps should be set to zero:

"training": {
"numb_steps": 0
}


Then, one can initialize the training from the frozen model and freeze the new model at once:

dp train input.json --init-frz-model frozen_model.pb

Note: At this time, this feature is only supported by se_e2_a descriptor with set_davg_true enabled, or hybrid composed of the above descriptors.