4

tensorflow DMA

What does the Y Y or Y N mean?

How to set the Y and N by myself?

By the way, the machine shutdown when it print (on the error machine)

Y Y
Y Y

And it run successfully on two GPU when print (on another machine)

Y N
N Y

So I wonder if it is the problem.

EDIT:

I use the program below could make it shutdown.

import numpy as np
import tensorflow as tf

with tf.device('/gpu:0'):
  W = tf.Variable([.3], tf.float32)
  b = tf.Variable([-.3], tf.float32)

with tf.device('/gpu:1'):
  x = tf.placeholder(tf.float32)
  linear_model = W * x + b
  y = tf.placeholder(tf.float32)

loss = tf.reduce_sum(tf.square(linear_model - y)) # sum of the squares

optimizer = tf.train.GradientDescentOptimizer(0.01)
train = optimizer.minimize(loss)

x_train = [1,2,3,4]
y_train = [0,-1,-2,-3]

init = tf.global_variables_initializer()
sess = tf.Session()
sess.run(init) 
for i in range(1000):
  sess.run(train, {x:x_train, y:y_train})

# evaluate training accuracy
curr_W, curr_b, curr_loss  = sess.run([W, b, loss], {x:x_train, y:y_train})
print("W: %s b: %s loss: %s"%(curr_W, curr_b, curr_loss))
DunkOnly
  • 1,682
  • 4
  • 17
  • 39

1 Answers1

0

It is Device interconnection. That shows how fast the data can be transferred between devices during multi-GPU training.

It is not a problem. The "NY" configuration depends on your hardware configuration. You cannot set it manually.

This post by @McAngus has a full answer.

Eugene
  • 130
  • 3
  • 17