Open convnets opened 6 years ago
In capsule_em.py,
capsule_em.py
votes = tf.reshape(tf.matmul(output, w), [batch_size, caps_num_i, caps_num_c, 16]) when training withtf.matmul, gpu usage is very low, usually around 50% or less sometimes. Why does this happen ?
votes = tf.reshape(tf.matmul(output, w), [batch_size, caps_num_i, caps_num_c, 16])
tf.matmul
In
capsule_em.py
,votes = tf.reshape(tf.matmul(output, w), [batch_size, caps_num_i, caps_num_c, 16])
when training withtf.matmul
, gpu usage is very low, usually around 50% or less sometimes. Why does this happen ?