Konquer your Mind

About how you preprocess the image

1

I saw you preprocess input image by subtracting the mean intensity of it rather than subtracting the mean intensity of the whole training images as people usually does. What is...

Is the left player or right player controlled by dqn? I can't find clue from your code.

a problem occurs at training model Tacotran-2

10

when the wavenet start to train after tacotron, python raise an error message that >AttributeError: type object 'Wrapper' has no attribute '_track_checkpointable' I print all members of tf.keras.layers.Wrapper with ```python...

there is no f distribution

3

f distribution which is frequently used in f test is not available in tfp.distributions. please support this distribution.

split2d operator seems only truncate a tensor half of its channels. the dY/dX = diag(I,0) rather then the objective returned by function split2d. could you explain why do you calculate...

I use AddOptimizerOps to create optimizer operators automatically. The optimizer operators are created even for blobs which are before the stopgradient operator. The stopgradient operator can surely prevent gradient calculation...

how to decide whether an operator is trainable

1

I am confusing about whether I shall put an operator into trainable_ops or non_trainable_ops. How do you decide it? I saw Scale operator was put in non_trainable_ops. Shall I put...

StopGradient operator doesn't stop AddGradientOps from creating gradient operators

7

ModelUtil::AddGradientOps function is supposed to stop creating gradient operators for operators before StopGradient operator. But it doesn't.

I read your python code and found out the decoding process is done by generating the face image directly from the code through the decoding network rather than by sampling...

after capture rgb and depth with depth_filter.py, the rgb and depth are saved under directory output. when I execute mvgmvs.sh, the script queries images under directory images. shall I just...