def save_model_bin(self, binFilePath, modelPath=None):
from fxpmath import Fxp
if modelPath != None:
self.model = load_model(modelPath, compile="False")
with open(binFilePath, 'wb') as binFile:
largest_inaccuracy = 0
for layer in self.model.layers:
g = layer.get_config()
h = layer.get_weights()
print(g)
#binFile.write(json.dumps(g).encode(encoding="ascii",errors="unknown char"))
# embedding = 1 * 68 * 8
# simple rnn = 8 * 8, 8 * 8, 8 * 1
# drop out: none
# dense: 8 * 1, 1 * 1
# activation: sigmoid
for i in h:
i = np.array(i)
for index, x in np.ndenumerate(i):
print(x)
h_fxp = Fxp(x, signed=True, n_word=16, n_frac=8)
difference = abs(h_fxp.get_val() - x)
if difference > largest_inaccuracy:
largest_inaccuracy = difference
print(h_fxp.bin())
binFile.write(h_fxp.bin().encode(
encoding="ascii", errors="unknown char"))
print("largest difference")
print(str(largest_inaccuracy))
def save_model_txt_binary(self, txtPath, modelPath=None):
from fxpmath import Fxp
import math
import json
if modelPath != None:
self.model = load_model(modelPath, compile="False")
with open(txtPath, 'w') as txtFile:
for layer in self.model.layers:
g = layer.get_config()
h = layer.get_weights()
txtFile.write(json.dumps(g))
txtFile.write("\n")
for i in h:
i = np.array(i)
for index, x in np.ndenumerate(i):
if g["name"] == "dropout":
continue
#if g["name"] == "embedding":
# print(index)
# row = math.floor(index / 4)
# col = index % 4
# txtFile.write("row:"+str(row)+"col:"+col)
# else:
# row = math.floor(index / 32)
# col = index % 32
# txtFile.write("row:"+str(row)+"col:"+col)
if len(index) > 1:
row = index[0]
col = index[1]
txtFile.write("row:" + str(row) + "col:" +
str(col))
else:
row = index[0]
txtFile.write("row:" + str(row))
h_fxp = Fxp(x, signed=True, n_word=16, n_frac=8)
txtFile.write("val:" + h_fxp.bin())
txtFile.write("\n")
word_bits = 16
frac_bits = 11
error = []
for f_ in files:
with open(f_, 'r') as handle:
data = np.genfromtxt(handle, delimiter=',')
fxp_val = np.zeros_like(data)
fxp_bin = np.ndarray(data.shape, dtype='U16')
for line in range(len(data)):
if "bias" in f_:
fxp_sample = Fxp(data[line], True, word_bits, frac_bits)
fxp_val[line] = fxp_sample.get_val()
fxp_bin[line] = fxp_sample.bin()
error.append(
np.nan_to_num((data[line] - fxp_val[line]) / data[line]))
else:
for column in range(len(data[line])):
fxp_sample = Fxp(data[line][column], True, word_bits,
frac_bits)
fxp_val[line][column] = fxp_sample.get_val()
fxp_bin[line][column] = fxp_sample.bin()
error.append(
np.nan_to_num(
(data[line][column] - fxp_val[line][column]) /
data[line][column]))
with open('report.txt', 'a') as r:
r.write(f_ + ' min: ' + str(np.min(fxp_val)) + ' - max: ' +
str(np.max(fxp_val)) + '\n')
import sys
sys.path.insert(1, '/Users/jingyuan/Desktop/dga/dga_detection_rnn')
from fxpmath import Fxp
import numpy as np
tanh_table_path = "conf/tanh_table.bin"
with open(tanh_table_path, 'wb') as binFile:
for entry in range(10):
xVal = -4 + entry * 8 / 9
print(xVal)
x_fxp = Fxp(xVal, signed=True, n_word=16, n_frac=12)
print(x_fxp.bin())
xTanh = np.tanh(xVal)
print(xTanh)
xTanh_fxp = Fxp(xTanh, signed=True, n_word=16, n_frac=12)
print(xTanh_fxp.bin())
binFile.write(x_fxp.bin().encode(encoding="ascii",
errors="unknown char"))
binFile.write(xTanh_fxp.bin().encode(encoding="ascii",
errors="unknown char"))