What is normal truncated distribution in TF learn explain

Recipe Objective

This recipe explains what is truncated normal distribution in TFLayer.

FastText and Word2Vec Word Embeddings Python Implementation

Truncated Normal

Its syntax is: tflearn.initializations.truncated_normal (shape=None, mean=0.0, stddev=0.02, dtype=tf.float32, seed=None)
where its arguments are shape which is a shape to initialize a Tensor, mean which is the mean of the truncated normal distribution, stddev which is the standard deviation of the truncated normal distribution, dtype which is the tensor data type and seed which is used to create a random seed for the distribution.

def Truncated_Normal(s=None, m=0.0, std=0.04, d=tf.float32, seed=None):

if shape:
    return tf.truncated_normal(s, mean=m, stddev=std, seed=seed, dtype=d)
else:
    return tf.truncated_normal_initializer(mean=m, stddev=std, seed=seed, dtype=d)