|View source on GitHub|
Initializer capable of adapting its scale to the shape of weights tensors.
tf.compat.v1.keras.initializers.lecun_normal( seed=None )
distribution="truncated_normal" or "untruncated_normal",
samples are drawn from a truncated/untruncated normal
distribution with a mean of zero and a standard deviation (after truncation,
stddev = sqrt(scale / n)
where n is:
- number of input units in the weight tensor, if mode = "fan_in"
- number of output units, if mode = "fan_out"
- average of the numbers of input and output units, if mode = "fan_avg"
distribution="uniform", samples are drawn from a uniform distribution
within [-limit, limit], with
limit = sqrt(3 * scale / n).
||Scaling factor (positive float).|
||One of "fan_in", "fan_out", "fan_avg".|
||Random distribution to use. One of "normal", "uniform".|
A Python integer. Used to create random seeds. See
Default data type, used if no