TensorFlow > sine curveの学習 > weightとbiasをnumpyファイル出力

動作環境

GeForce GTX 1070 (8GB)
ASRock Z170M Pro4S [Intel Z170chipset]
Ubuntu 14.04 LTS desktop amd64
TensorFlow v0.11
cuDNN v5.1 for Linux
CUDA v8.0
Python 2.7.6
IPython 5.1.0 -- An enhanced Interactive Python.

関連　http://qiita.com/7of9/items/b364d897b95476a30754

sine curveを学習した時のweightとbiasをもとに自分でネットワークを再現して出力を計算しようとしている。

np.save()を使うことになるだろうか。
http://qiita.com/7of9/items/c730990479687ec2e959

input.csv生成

書出しcode

linreg2_reprod.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import sys
import tensorflow as tf
import tensorflow.contrib.slim as slim
import numpy as np

# ファイル名の Queue を作成
filename_queue = tf.train.string_input_producer(["input.csv"])

# CSV を parse
reader = tf.TextLineReader()
key, value = reader.read(filename_queue)
input1, output = tf.decode_csv(value, record_defaults=[[0.], [0.]])
inputs = tf.pack([input1])
output = tf.pack([output])

batch_size=4 # [4]
inputs_batch, output_batch = tf.train.shuffle_batch([inputs, output], batch_size, capacity=40, min_after_dequeue=batch_size)

input_ph = tf.placeholder("float", [None,1])
output_ph = tf.placeholder("float",[None,1])

## NN のグラフ生成
hiddens = slim.stack(input_ph, slim.fully_connected, [1,7,7,7], 
  activation_fn=tf.nn.sigmoid, scope="hidden")
prediction = slim.fully_connected(hiddens, 1, activation_fn=tf.nn.sigmoid, scope="output")
loss = tf.contrib.losses.mean_squared_error(prediction, output_ph)

train_op = slim.learning.create_train_op(loss, tf.train.AdamOptimizer(0.001))

init_op = tf.initialize_all_variables()

with tf.Session() as sess:
  coord = tf.train.Coordinator()
  threads = tf.train.start_queue_runners(sess=sess, coord=coord)

  try:
    sess.run(init_op)
    for i in range(30000): #[10000]
      inpbt, outbt = sess.run([inputs_batch, output_batch])
      _, t_loss = sess.run([train_op, loss], feed_dict={input_ph:inpbt, output_ph: outbt})

      if (i+1) % 100 == 0:
        print("%d,%f" % (i+1, t_loss))

    # output to npy 
    model_variables = slim.get_model_variables()
    res = sess.run(model_variables)
    np.save('model_variables.npy', res)

  finally:
    coord.request_stop()

  coord.join(threads)

実行するとmodel_variables.npyというファイルが生成された。

読込みcode

read_model_var.py

import numpy as np

model_var = np.load('model_variables.npy')
print (model_var)

実行

$python read_model_var.py
[array([[-5.22224426]], dtype=float32) array([ 1.78673065], dtype=float32)
 array([[-4.58573723, -4.27450418, -3.75889063,  4.51949883, -4.02780342,
        -4.10122681,  4.0842309 ]], dtype=float32)
 array([ 0.97008353,  0.70625514,  0.27048966, -0.83405548,  0.57475132,
        0.64893931, -0.45576799], dtype=float32)
 array([[ 1.84565568,  1.49943328,  1.65000439,  2.44512415,  2.09082389,
         2.37278032,  1.65048397],
       [ 1.4404825 ,  0.94583297,  1.89746273,  2.51769876,  1.5209198 ,
         1.94879484,  1.19026875],
       [ 1.49121964,  1.18247831,  1.51108956,  1.18311214,  0.88615912,
         1.1377635 ,  1.33165669],
       [-2.37849569,  0.55434752, -2.70425415, -2.46672988, -2.60600066,
        -2.62578273, -2.78632236],
       [ 0.82718104,  1.42030048,  1.06626236,  1.53540218,  1.55356538,
         1.84659779,  1.25057125],
       [ 1.30038536,  1.45610416,  1.77369738,  2.22379041,  2.24454832,
         2.2828269 ,  1.8470453 ],
       [-2.0469408 ,  1.2918936 , -1.87940514, -2.31857991, -2.2989893 ,
        -2.25665474, -1.73469198]], dtype=float32)
 array([-1.1732285 ,  0.97712201, -1.06408942, -0.75335336, -0.74013597,
       -0.65020418, -1.12149072], dtype=float32)
 array([[ 0.9464646 , -0.11631355, -0.16895044,  0.36979192,  1.05458641,
         0.76118785, -0.4746716 ],
       [-1.47142816, -0.67341083,  0.86623627, -1.63780856, -1.4503684 ,
        -0.71064085,  1.02097607],
       [ 1.58198178,  1.17400146, -1.40148842,  0.5268032 ,  1.45251906,
         0.28084767, -0.8032999 ],
       [ 1.61421323,  1.7214433 , -1.26047742,  1.9050349 ,  1.21235812,
         1.24344146, -2.05446649],
       [ 0.47106522,  1.23160982, -1.13437021,  1.25667596,  0.68895262,
         1.86171079, -0.76870179],
       [ 1.17536426,  1.74713993, -1.9202348 ,  1.397084  ,  1.61537564,
         2.0741148 , -1.37128556],
       [ 0.85221511,  0.6925481 , -0.62384129,  1.12779391,  0.73884082,
         0.08157811, -1.01910996]], dtype=float32)
 array([-1.03282344, -1.26972938,  1.24813604, -1.1338433 , -1.21751046,
       -0.94728774,  1.32416117], dtype=float32)
 array([[-2.49866319],
       [-1.76430416],
       [ 1.60942447],
       [-2.18569684],
       [-2.33169866],
       [-1.27172542],
       [ 2.39537191]], dtype=float32)
 array([-0.45804733], dtype=float32)]

読めていそうだ。

これで学習を繰り返さずに、学習結果を使った処理を繰り返し試すことができる。