Python get_ds_simpleの例

コード例 #1

ファイル: test_image.py プロジェクト: undertherain/dagen

 def test_merge_w_channels(self):
     X_train, Y_train = get_ds_simple(cnt_samples=10)
     X_train = np.expand_dims(X_train, axis=1).astype(np.float32) / 255
     Y_train = Y_train[:, np.newaxis]
     print(X_train.shape, Y_train.shape)
     im = merge_samples(X_train, Y_train)
     im.save("/tmp/simple_channel.png")

コード例 #2

ファイル: test_image.py プロジェクト: undertherain/dagen

 def test_size(self):
     X_train, Y_train = get_ds_simple(dim_image=128, cnt_samples=10)
     print(X_train.shape, Y_train.shape)
     im = merge_samples(X_train, Y_train)
     im.save("/tmp/size128.png")

コード例 #3

ファイル: classify.py プロジェクト: undertherain/nuts-and-bolts

import numpy as np
import chainer
import chainer.functions as F
import chainer.links as L

import dagen
import dagen.image
from dagen.image.image import get_ds_simple

from ..trainer import train

params = {}
params["batch_size"] = 8

X_train, Y_train = get_ds_simple(cnt_samples=1000)
X_test, Y_test = get_ds_simple(cnt_samples=100)
X_train = np.expand_dims(X_train, axis=1).astype(np.float32) / 255
Y_train = Y_train[:, np.newaxis]
X_test = np.expand_dims(X_test, axis=1).astype(np.float32) / 255
Y_test = Y_test[:, np.newaxis]
print(X_train.shape)
print(Y_train.shape)


class Net(chainer.Chain):

    def __init__(self, train=True):
        super(Net, self).__init__(
            conv1=L.Convolution2D(1, 16, 2),
            conv2=L.Convolution2D(None, 16, 2),
            l1=L.Linear(None, 10),

コード例 #4

ファイル: test_image.py プロジェクト: undertherain/dagen

 def test_simple(self):
     X_train, Y_train = get_ds_simple(cnt_samples=10)
     print(X_train.shape, Y_train.shape)
     im = merge_samples(X_train, Y_train)
     im.save("/tmp/simple.png")

コード例 #5

ファイル: autoencoder.py プロジェクト: undertherain/nuts-and-bolts

import chainer
import chainer.functions as F
import chainer.links as L

import dagen
import dagen.image
from dagen.image.image import get_ds_simple
from dagen.image.image import merge_samples

from ..trainer import train

params = {"nb_epoch": 10}
params["batch_size"] = 8
params["gpus"] = [0]

X_train, Y_train = get_ds_simple(dim_image=32, cnt_samples=1000)
X_train = np.expand_dims(X_train, axis=1).astype(np.float32) / 255
print(X_train.shape)


class Net(chainer.Chain):
    def __init__(self, train=True):
        super(Net, self).__init__(
            conv_e_1=L.Convolution2D(None, 32, 3, pad=1),
            conv_e_2=L.Convolution2D(None, 32, 3, pad=1),
            conv_e_3=L.Convolution2D(None, 32, 3, pad=1),
            conv_d_1=L.Convolution2D(None, 32, 3, pad=1),
            conv_d_2=L.Convolution2D(None, 32, 3, pad=1),
            conv_d_3=L.Convolution2D(None, 32, 3, pad=1),
            conv_d_4=L.Convolution2D(None, 1, 3, pad=1),
            # dc1 = L.Deconvolution2D(in_channels=None, out_channels=32, ksize=2, stride=2, pad=0 ),