Related questions

What does the tf_exported_symbols.lds do

2018-01-11 20:18 superd imported from Stackoverflow

I'm learning bazel and tensorflow, and when I check the tensorflow code I find this file: https://github.com/tensorflow/tensorflow/blob/master/tensorflow/tf_version_script.lds

So what does this file actually do?

See also questions close to this topic

  • How can I call up the weight and bias from the model?

    I'd like to ask you about the model Tensorflow. Is there a way to designate an intrinsic value when training other models by calling up the weight, bias of the stored model?

  • Can change network architecture during training in tensorflow?

    I want to change channel dimensions of convolution layer during training, but in testing, always keeping same dimensions. I tried to implement this in tensorflow, and i failed. If the output dimension changes during training, an error occurs because the tensor flow does not recognize the changed graph. Does not the tensor flow implement these features? If not, is there any other way to implement this feature?

  • Pandas Data Normalization is giving weird output Python 3

    I have tried to create a train and test dataset and have found that the output after training what a weird graph which was not expected.

    The issue I came across is because of the use of the dataset normalization. The output of the training dataset and testing dataset is somewhat weird.

    Have a look at the code: 

    dataset_train = pd.read_csv('TrainEURUSD.csv',usecols=[1,2,3,4,5,6,7,8,9,10,11,12,13])
dataset_train.head(10)

    Output: 

    0   1.16833 1.16835 1.16823 1.16831 1.16833 66.101695   84.903097   0.000164    -6.314161e-06   84.903097   1.168795    1.168415    1.168034
1   1.16827 1.16835 1.16827 1.16833 1.16842 88.607595   90.104971   0.000150    -9.244212e-07   90.104971   1.168840    1.168437    1.168033
2   1.16842 1.16843 1.16841 1.16842 1.16842 100.000000  75.569106   0.000164    3.503926e-06    75.569106   1.168882    1.168458    1.168034
3   1.16833 1.16845 1.16823 1.16842 1.16833 81.707317   51.326681   0.000156    -1.376676e-05   51.326681   1.168885    1.168462    1.168038
4   1.16831 1.16835 1.16831 1.16833 1.16825 45.000000   28.316261   0.000162    -4.390005e-05   28.316261   1.168889    1.168466    1.168042
5   1.16813 1.16825 1.16813 1.16825 1.16811 27.272727   27.675236   0.000168    -6.926663e-05   27.675236   1.168890    1.168471    1.168052
6   1.16828 1.16828 1.16811 1.16811 1.16828 12.676056   28.729254   0.000159    -8.585350e-05   28.729254   1.168898    1.168488    1.168078
7   1.16824 1.16832 1.16814 1.16828 1.16823 43.076923   35.059457   0.000158    -6.783461e-05   35.059457   1.168882    1.168511    1.168139
8   1.16840 1.16840 1.16820 1.16823 1.16840 30.434783   24.246582   0.000157    -7.448502e-05   24.246582   1.168878    1.168521    1.168165
9   1.16821 1.16840 1.16821 1.16840 1.16821 31.666667   18.546099   0.000166    -6.451244e-05   18.546099   1.168868    1.168538    1.168207

    Next: 

    dataset_test = pd.read_csv('Test_EURUSD.csv',usecols=[1,2,3,4,5,6,7,8,9,10,11,12,13])
dataset_test.head(10)

    Output: 

    0   1.16610 1.16615 1.16601 1.16603 1.16610 25.000000   53.309179   0.000080    -0.000027   53.309179   1.166276    1.166168    1.166061
1   1.16623 1.16623 1.16610 1.16610 1.16623 56.666667   63.494364   0.000074    -0.000010   63.494364   1.166274    1.166180    1.166086
2   1.16617 1.16624 1.16617 1.16623 1.16618 78.260870   52.012882   0.000071    0.000003    52.012882   1.166290    1.166190    1.166090
3   1.16613 1.16618 1.16612 1.16618 1.16613 55.555556   34.259259   0.000068    -0.000009   34.259259   1.166301    1.166193    1.166085
4   1.16615 1.16615 1.16612 1.16613 1.16615 22.222222   23.518519   0.000067    -0.000018   23.518519   1.166312    1.166198    1.166084
5   1.16613 1.16615 1.16613 1.16615 1.16612 25.000000   24.005848   0.000069    -0.000017   24.005848   1.166334    1.166208    1.166083
6   1.16617 1.16621 1.16612 1.16612 1.16617 23.333333   25.672515   0.000073    -0.000019   25.672515   1.166351    1.166217    1.166084
7   1.16618 1.16626 1.16615 1.16617 1.16618 23.684211   35.355054   0.000069    -0.000013   35.355054   1.166388    1.166233    1.166078
8   1.16614 1.16619 1.16614 1.16618 1.16614 30.000000   53.956044   0.000074    -0.000014   53.956044   1.166420    1.166246    1.166072
9   1.16618 1.16618 1.16609 1.16614 1.16618 52.380952   68.465848   0.000071    -0.000017   68.465848   1.166447    1.166259    1.166070

    Then I normalized the datasets, using the following: 

    datasetTrain = (dataset_train - dataset_train.mean()) / (dataset_train.max() - dataset_train.min())
datasetTrain.head(10)

    Output: 

    0   0.019674    0.017293    0.019989    0.019084    0.019678    0.160252    0.348290    -0.010598   -0.002514   0.348290    0.020240    0.023543    0.025458
1   0.017889    0.017293    0.021206    0.019679    0.022357    0.385311    0.400309    -0.019556   -0.000335   0.400309    0.021616    0.024238    0.025413
2   0.022352    0.019660    0.025462    0.022357    0.022357    0.499235    0.254950    -0.010598   0.001454    0.254950    0.022881    0.024918    0.025449
3   0.019674    0.020251    0.019989    0.022357    0.019678    0.316308    0.012526    -0.015313   -0.005526   0.012526    0.022982    0.025028    0.025563
4   0.019079    0.017293    0.022422    0.019679    0.017297    -0.050765   -0.217578   -0.011541   -0.017706   -0.217578   0.023099    0.025154    0.025691
5   0.013723    0.014336    0.016949    0.017298    0.013131    -0.228038   -0.223988   -0.007769   -0.027959   -0.223988   0.023134    0.025328    0.025996
6   0.018186    0.015223    0.016341    0.013131    0.018190    -0.374004   -0.213448   -0.013899   -0.034664   -0.213448   0.023353    0.025865    0.026824
7   0.016996    0.016406    0.017253    0.018191    0.016702    -0.069996   -0.150146   -0.014370   -0.027381   -0.150146   0.022889    0.026576    0.028693
8   0.021756    0.018772    0.019077    0.016703    0.021761    -0.196417   -0.258275   -0.014842   -0.030069   -0.258275   0.022761    0.026923    0.029506
9   0.016103    0.018772    0.019381    0.021762    0.016107    -0.184098   -0.315280   -0.008712   -0.026038   -0.315280   0.022459    0.027429    0.030806

    And for the test dataset: 

    datasetTest = (dataset_test - dataset_test.mean()) / (dataset_test.max() - dataset_test.min())
datasetTest.head(10)

    Output: 

    0   -0.148985   -0.151914   -0.150167   -0.152497   -0.148978   -0.248599   0.034471    -0.083573   -0.022027   0.034471    -0.167361   -0.152037   -0.135121
1   -0.142436   -0.147849   -0.145610   -0.148971   -0.142429   0.068068    0.136323    -0.090135   -0.007864   0.136323    -0.167459   -0.151432   -0.133844
2   -0.145459   -0.147341   -0.142066   -0.142422   -0.144947   0.284010    0.021508    -0.094237   0.002898    0.021508    -0.166600   -0.150906   -0.133647
3   -0.147474   -0.150390   -0.144598   -0.144941   -0.147466   0.056957    -0.156028   -0.097518   -0.007581   -0.156028   -0.166023   -0.150748   -0.133897
4   -0.146466   -0.151914   -0.144598   -0.147460   -0.146459   -0.276377   -0.263436   -0.098339   -0.014414   -0.263436   -0.165428   -0.150485   -0.133959
5   -0.147474   -0.151914   -0.144091   -0.146452   -0.147970   -0.248599   -0.258562   -0.095878   -0.013806   -0.258562   -0.164281   -0.149933   -0.133990
6   -0.145459   -0.148866   -0.144598   -0.147963   -0.145451   -0.265266   -0.241896   -0.091776   -0.015730   -0.241896   -0.163404   -0.149459   -0.133914
7   -0.144955   -0.146325   -0.143079   -0.145445   -0.144947   -0.261757   -0.145070   -0.095878   -0.010578   -0.145070   -0.161441   -0.148644   -0.134220
8   -0.146970   -0.149882   -0.143585   -0.144941   -0.146963   -0.198599   0.040939    -0.090956   -0.011433   0.040939    -0.159733   -0.147960   -0.134537
9   -0.144955   -0.150390   -0.146117   -0.146956   -0.144947   0.025211    0.186038    -0.093417   -0.014001   0.186038    -0.158295   -0.147302   -0.134643

    As one can see the difference between the train dataset and test dataset normalized, the values corresponding to the columns are reverse in sign. That is the value is positive in the train dataset column whereas in test is negative.

    Here are the link for the test and train dataset.

    Kindly, suggest me what I have missed. Or am I making some mistake.