deepmd.nvnmd.utils package

class deepmd.nvnmd.utils.Encode[source]

Bases: object

Encoding value as hex, bin, and dec format

Methods

`bin2hex`(data)	Convert binary string list to hex string list
`bin2hex_str`(sbin)	Convert binary string to hex string
`check_dec`(idec, nbit[, signed, name])	Check whether the data (idec) is in the range range is \([0, 2^nbit-1]\) for unsigned range is \([-2^{nbit-1}, 2^{nbit-1}-1]\) for signed
`dec2bin`(idec[, nbit, signed, name])	Convert dec array to binary string list
`extend_bin`(slbin, nfull)	Extend the element of list (slbin) to the length (nfull)
`extend_hex`(slhex, nfull)	Extend the element of list (slhex) to the length (nfull)
`extend_list`(slbin, nfull)	Extend the list (slbin) to the length (nfull) the attched element of list is 0
`hex2bin`(data)	Convert hex string list to binary string list
`hex2bin_str`(shex)	Convert hex string to binary string
`merge_bin`(slbin, nmerge)	Merge binary string list per nmerge value
`qc`(v[, nbit])	Quantize value using ceil
`qf`(v[, nbit])	Quantize value using floor
`qr`(v[, nbit])	Quantize value using round
`reverse_bin`(slbin, nreverse)	Reverse binary string list per nreverse value
`split_bin`(sbin, nbit)	Split sbin into many segment with the length nbit

bin2hex(data)[source]: Convert binary string list to hex string list

bin2hex_str(sbin)[source]: Convert binary string to hex string

check_dec(idec, nbit, signed=False, name='')[source]: Check whether the data (idec) is in the range range is \([0, 2^nbit-1]\) for unsigned range is \([-2^{nbit-1}, 2^{nbit-1}-1]\) for signed

dec2bin(idec, nbit=10, signed=False, name='')[source]: Convert dec array to binary string list

extend_bin(slbin, nfull)[source]

Extend the element of list (slbin) to the length (nfull)

such as, when

slbin = [‘10010’,’10100’],

nfull = 6

extent to

[‘010010’,’010100’]

extend_hex(slhex, nfull)[source]: Extend the element of list (slhex) to the length (nfull)

extend_list(slbin, nfull)[source]

Extend the list (slbin) to the length (nfull) the attched element of list is 0

such as, when

slbin = [‘10010’,’10100’],

nfull = 4

extent it to

[‘10010’,’10100’,’00000’,’00000]

hex2bin(data)[source]: Convert hex string list to binary string list

hex2bin_str(shex)[source]: Convert hex string to binary string

merge_bin(slbin, nmerge)[source]: Merge binary string list per nmerge value

qc(v, nbit: int = 14)[source]: Quantize value using ceil

qf(v, nbit: int = 14)[source]: Quantize value using floor

qr(v, nbit: int = 14)[source]: Quantize value using round

reverse_bin(slbin, nreverse)[source]: Reverse binary string list per nreverse value

split_bin(sbin, nbit: int)[source]: Split sbin into many segment with the length nbit

class deepmd.nvnmd.utils.FioBin[source]

Bases: object

Input and output for binary file

Methods

`load`([file_name, default_value])	Load binary file into bytes value
`save`([file_name, data])	Save hex string into binary file

load(file_name='', default_value='')[source]: Load binary file into bytes value

save(file_name: str = '', data: str = '')[source]: Save hex string into binary file

class deepmd.nvnmd.utils.FioDic[source]

Bases: object

Input and output for dict class data the file can be .json or .npy file containing a dictionary

Methods

update(jdata, jdata_o)

Update key-value pair is key in jdata_o.keys()

get
load
save

get(jdata, key, default_value)[source]

load(file_name='', default_value={})[source]

save(file_name='', dic={})[source]

update(jdata, jdata_o)[source]: Update key-value pair is key in jdata_o.keys()

class deepmd.nvnmd.utils.FioTxt[source]

Bases: object

Input and output for .txt file with string

Methods

`load`([file_name, default_value])	Load .txt file into string list
`save`([file_name, data])	Save string list into .txt file

load(file_name='', default_value=[])[source]: Load .txt file into string list

save(file_name: str = '', data: list = [])[source]: Save string list into .txt file

deepmd.nvnmd.utils.get_filter_weight(weights: dict, spe_i: int, spe_j: int, layer_l: int)[source]

Get weight and bias of embedding network

Parameters

spe_i(int): special order of central atom i 0~ntype-1
spe_j(int): special order of neighbor atom j 0~ntype-1
layer_l: layer order in embedding network 1~nlayer

deepmd.nvnmd.utils.get_fitnet_weight(weights: dict, spe_i: int, layer_l: int, nlayer: int = 10)[source]

Get weight and bias of fitting network

Parameters

spe_i(int): special order of central atom i 0~ntype-1
layer_l(int): layer order in embedding network 0~nlayer-1

deepmd.nvnmd.utils.map_nvnmd(x, map_y, map_dy, prec, nbit=None)[source]: Mapping function implemented by numpy

deepmd.nvnmd.utils.nvnmd_args()[source]

deepmd.nvnmd.utils.one_layer(inputs, outputs_size, activation_fn=<function tanh>, precision=tf.float64, stddev=1.0, bavg=0.0, name='linear', reuse=None, seed=None, use_timestep=False, trainable=True, useBN=False, uniform_seed=False, initial_variables=None, mixed_prec=None, final_layer=False)[source]: Build one layer with continuous or quantized value. Its weight and bias can be initialed with random or constant value.

Submodules

deepmd.nvnmd.utils.argcheck module

deepmd.nvnmd.utils.argcheck.nvnmd_args()[source]

deepmd.nvnmd.utils.config module

class deepmd.nvnmd.utils.config.NvnmdConfig(jdata: dict)[source]

Bases: object

Configuration for NVNMD record the message of model such as size, using nvnmd or not

Parameters

jdata: a dictionary of input script

References

DOI: 10.1038/s41524-022-00773-z

Methods

`disp_message`()	Display the log of NVNMD
`get_deepmd_jdata`()	Generate input script with member element one by one
`get_dscp_jdata`()	Generate model/descriptor in input script
`get_fitn_jdata`()	Generate model/fitting_net in input script
`get_learning_rate_jdata`()	Generate learning_rate in input script
`get_loss_jdata`()	Generate loss in input script
`get_model_jdata`()	Generate model in input script
`get_nvnmd_jdata`()	Generate nvnmd in input script
`get_training_jdata`()	Generate training in input script
`init_ctrl`(jdata[, jdata_parent])	Initial members about control signal
`init_dscp`(jdata[, jdata_parent])	Initial members about descriptor
`init_fitn`(jdata[, jdata_parent])	Initial members about fitting network
`init_from_config`(jdata)	Initial member element one by one
`init_from_deepmd_input`(jdata)	Initial members with input script of deepmd
`init_from_jdata`([jdata])	Initial this class with jdata loaded from input script
`init_nbit`(jdata[, jdata_parent])	Initial members about quantification precision
`init_net_size`()	Initial net_size
`init_size`(jdata[, jdata_parent])	Initial members about ram capacity
`init_train_mode`([mod])	Configure for taining cnn or qnn
`init_value`()	Initial member with dict
`save`([file_name])	Save all configuration to file

disp_message()[source]: Display the log of NVNMD

get_deepmd_jdata()[source]: Generate input script with member element one by one

get_dscp_jdata()[source]: Generate model/descriptor in input script

get_fitn_jdata()[source]: Generate model/fitting_net in input script

get_learning_rate_jdata()[source]: Generate learning_rate in input script

get_loss_jdata()[source]: Generate loss in input script

get_model_jdata()[source]: Generate model in input script

get_nvnmd_jdata()[source]: Generate nvnmd in input script

get_training_jdata()[source]: Generate training in input script

init_ctrl(jdata: dict, jdata_parent: dict = {}) → dict[source]: Initial members about control signal

init_dscp(jdata: dict, jdata_parent: dict = {}) → dict[source]: Initial members about descriptor

init_fitn(jdata: dict, jdata_parent: dict = {}) → dict[source]: Initial members about fitting network

init_from_config(jdata)[source]: Initial member element one by one

init_from_deepmd_input(jdata)[source]: Initial members with input script of deepmd

init_from_jdata(jdata: dict = {})[source]: Initial this class with jdata loaded from input script

init_nbit(jdata: dict, jdata_parent: dict = {}) → dict[source]: Initial members about quantification precision

init_net_size()[source]: Initial net_size

init_size(jdata: dict, jdata_parent: dict = {}) → dict[source]: Initial members about ram capacity

init_train_mode(mod='cnn')[source]: Configure for taining cnn or qnn

init_value()[source]: Initial member with dict

save(file_name=None)[source]: Save all configuration to file

deepmd.nvnmd.utils.encode module

class deepmd.nvnmd.utils.encode.Encode[source]

Bases: object

Encoding value as hex, bin, and dec format

Methods

`bin2hex`(data)	Convert binary string list to hex string list
`bin2hex_str`(sbin)	Convert binary string to hex string
`check_dec`(idec, nbit[, signed, name])	Check whether the data (idec) is in the range range is \([0, 2^nbit-1]\) for unsigned range is \([-2^{nbit-1}, 2^{nbit-1}-1]\) for signed
`dec2bin`(idec[, nbit, signed, name])	Convert dec array to binary string list
`extend_bin`(slbin, nfull)	Extend the element of list (slbin) to the length (nfull)
`extend_hex`(slhex, nfull)	Extend the element of list (slhex) to the length (nfull)
`extend_list`(slbin, nfull)	Extend the list (slbin) to the length (nfull) the attched element of list is 0
`hex2bin`(data)	Convert hex string list to binary string list
`hex2bin_str`(shex)	Convert hex string to binary string
`merge_bin`(slbin, nmerge)	Merge binary string list per nmerge value
`qc`(v[, nbit])	Quantize value using ceil
`qf`(v[, nbit])	Quantize value using floor
`qr`(v[, nbit])	Quantize value using round
`reverse_bin`(slbin, nreverse)	Reverse binary string list per nreverse value
`split_bin`(sbin, nbit)	Split sbin into many segment with the length nbit

bin2hex(data)[source]: Convert binary string list to hex string list

bin2hex_str(sbin)[source]: Convert binary string to hex string

check_dec(idec, nbit, signed=False, name='')[source]: Check whether the data (idec) is in the range range is \([0, 2^nbit-1]\) for unsigned range is \([-2^{nbit-1}, 2^{nbit-1}-1]\) for signed

dec2bin(idec, nbit=10, signed=False, name='')[source]: Convert dec array to binary string list

extend_bin(slbin, nfull)[source]

Extend the element of list (slbin) to the length (nfull)

such as, when

slbin = [‘10010’,’10100’],

nfull = 6

extent to

[‘010010’,’010100’]

extend_hex(slhex, nfull)[source]: Extend the element of list (slhex) to the length (nfull)

extend_list(slbin, nfull)[source]

Extend the list (slbin) to the length (nfull) the attched element of list is 0

such as, when

slbin = [‘10010’,’10100’],

nfull = 4

extent it to

[‘10010’,’10100’,’00000’,’00000]

hex2bin(data)[source]: Convert hex string list to binary string list

hex2bin_str(shex)[source]: Convert hex string to binary string

merge_bin(slbin, nmerge)[source]: Merge binary string list per nmerge value

qc(v, nbit: int = 14)[source]: Quantize value using ceil

qf(v, nbit: int = 14)[source]: Quantize value using floor

qr(v, nbit: int = 14)[source]: Quantize value using round

reverse_bin(slbin, nreverse)[source]: Reverse binary string list per nreverse value

split_bin(sbin, nbit: int)[source]: Split sbin into many segment with the length nbit

deepmd.nvnmd.utils.fio module

class deepmd.nvnmd.utils.fio.Fio[source]

Bases: object

Basic class for FIO

Methods

create_file_path
exits
get_file_list
is_file
is_path
mkdir

create_file_path(file_name='')[source]

exits(file_name='')[source]

get_file_list(path) → list[source]

is_file(file_name)[source]

is_path(path)[source]

mkdir(path_name='')[source]

class deepmd.nvnmd.utils.fio.FioBin[source]

Bases: object

Input and output for binary file

Methods

`load`([file_name, default_value])	Load binary file into bytes value
`save`([file_name, data])	Save hex string into binary file

load(file_name='', default_value='')[source]: Load binary file into bytes value

save(file_name: str = '', data: str = '')[source]: Save hex string into binary file

class deepmd.nvnmd.utils.fio.FioDic[source]

Bases: object

Input and output for dict class data the file can be .json or .npy file containing a dictionary

Methods

update(jdata, jdata_o)

Update key-value pair is key in jdata_o.keys()

get
load
save

get(jdata, key, default_value)[source]

load(file_name='', default_value={})[source]

save(file_name='', dic={})[source]

update(jdata, jdata_o)[source]: Update key-value pair is key in jdata_o.keys()

class deepmd.nvnmd.utils.fio.FioJsonDic[source]

Bases: object

Input and output for .json file containing dictionary

Methods

`load`([file_name, default_value])	Load .json file into dict
`save`([file_name, dic])	Save dict into .json file

load(file_name='', default_value={})[source]: Load .json file into dict

save(file_name='', dic={})[source]: Save dict into .json file

class deepmd.nvnmd.utils.fio.FioNpyDic[source]

Bases: object

Input and output for .npy file containing dictionary

Methods

load
save

load(file_name='', default_value={})[source]

save(file_name='', dic={})[source]

class deepmd.nvnmd.utils.fio.FioTxt[source]

Bases: object

Input and output for .txt file with string

Methods

`load`([file_name, default_value])	Load .txt file into string list
`save`([file_name, data])	Save string list into .txt file

load(file_name='', default_value=[])[source]: Load .txt file into string list

save(file_name: str = '', data: list = [])[source]: Save string list into .txt file

deepmd.nvnmd.utils.network module

deepmd.nvnmd.utils.network.get_sess()[source]

deepmd.nvnmd.utils.network.matmul2_qq(a, b, nbit)[source]: Quantized matmul operation for 2d tensor. a and b is input tensor, nbit represent quantification precision

deepmd.nvnmd.utils.network.matmul3_qq(a, b, nbit)[source]: Quantized matmul operation for 3d tensor. a and b is input tensor, nbit represent quantification precision

deepmd.nvnmd.utils.network.one_layer(inputs, outputs_size, activation_fn=<function tanh>, precision=tf.float64, stddev=1.0, bavg=0.0, name='linear', reuse=None, seed=None, use_timestep=False, trainable=True, useBN=False, uniform_seed=False, initial_variables=None, mixed_prec=None, final_layer=False)[source]: Build one layer with continuous or quantized value. Its weight and bias can be initialed with random or constant value.

deepmd.nvnmd.utils.network.one_layer_wb(shape, outputs_size, bavg, stddev, precision, trainable, initial_variables, seed, uniform_seed, name)[source]

deepmd.nvnmd.utils.network.qf(x, nbit)[source]: Quantize and floor tensor x with quantification precision nbit.

deepmd.nvnmd.utils.network.qr(x, nbit)[source]: Quantize and round tensor x with quantification precision nbit.

deepmd.nvnmd.utils.network.tanh2(x, nbit=- 1, nbit2=- 1)[source]: User-defined activation function tanh2

deepmd.nvnmd.utils.network.tanh4(x, nbit=- 1, nbit2=- 1)[source]: User-defined activation function tanh4

deepmd.nvnmd.utils.op module

deepmd.nvnmd.utils.op.map_nvnmd(x, map_y, map_dy, prec, nbit=None)[source]: Mapping function implemented by numpy

deepmd.nvnmd.utils.weight module

deepmd.nvnmd.utils.weight.get_constant_initializer(weights, name)[source]: Get initial value by name and create a initializer

deepmd.nvnmd.utils.weight.get_filter_weight(weights: dict, spe_i: int, spe_j: int, layer_l: int)[source]

Get weight and bias of embedding network

Parameters

spe_i(int): special order of central atom i 0~ntype-1
spe_j(int): special order of neighbor atom j 0~ntype-1
layer_l: layer order in embedding network 1~nlayer

deepmd.nvnmd.utils.weight.get_fitnet_weight(weights: dict, spe_i: int, layer_l: int, nlayer: int = 10)[source]

Get weight and bias of fitting network

Parameters

spe_i(int): special order of central atom i 0~ntype-1
layer_l(int): layer order in embedding network 0~nlayer-1

deepmd.nvnmd.utils.weight.get_normalize(weights: dict)[source]: Get normalize parameter (avg and std) of \(s_{ji}\)

deepmd.nvnmd.utils.weight.get_rng_s(weights: dict)[source]: Guess the range of \(s_{ji}\)

deepmd.nvnmd.utils.weight.get_weight(weights, key)[source]: Get weight value according to key