TensorBoard:如何绘制梯度的直方图? [英] TensorBoard: How to plot histogram for gradients?
问题描述
TensorBoard具有在会话期间绘制张量直方图的功能.我想要训练时渐变的直方图.
TensorBoard had the function to plot histograms of Tensors at session-time. I want a histogram for the gradients during training.
tf.gradients(yvars,xvars)返回一个渐变列表.
但是，tf.histogram_summary('name',Tensor)仅接受张量，而不接受张量列表.
However, tf.histogram_summary('name',Tensor) accepts only Tensors, not lists of Tensors.
暂时，我做了一个解决方法.我将所有张量展平为列向量，并将它们连接起来:
For the time being, I made a work-around. I flatten all Tensors to a column vector and concatenate them:
for l in xrange(listlength):
col_vec = tf.reshape(grads[l],[-1,1])
g = tf.concat(0,[g,col_vec])
grad_hist = tf.histogram_summary("name", g)
for l in xrange(listlength):
col_vec = tf.reshape(grads[l],[-1,1])
g = tf.concat(0,[g,col_vec])
grad_hist = tf.histogram_summary("name", g)
绘制梯度直方图的更好方法是什么?
What would be a better way to plot the histogram for the gradient?
这似乎很平常，所以我希望TensorFlow对此具有专用功能.
It seems a common thing to do, so I hope TensorFlow would have a dedicated function for this.
推荐答案
Another solution (based on this quora answer) is to access the gradients directly from the optimizer you are already using.
optimizer = tf.train.AdamOptimizer(..)
grads = optimizer.compute_gradients(loss)
grad_summ_op = tf.summary.merge([tf.summary.histogram("%s-grad" % g[1].name, g[0]) for g in grads])
grad_vals = sess.run(fetches=grad_summ_op, feed_dict = feed_dict)
writer['train'].add_summary(grad_vals)
这篇关于TensorBoard:如何绘制梯度的直方图?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
