Resolve "exponential decay learning rate is not working" (!9) · Merge requests · bob / bob.learn.tensorflow · GitLab

Snippets Groups Projects

This is an archived project. Repository and other project resources are read-only.

Merged Amir MOHAMMADI requested to merge 31-exponential-decay-learning-rate-is-not-working into master 7 years ago

1 unresolved thread

Closes #31 (closed)

Edited 7 years ago by Amir MOHAMMADI

Activity

Amir MOHAMMADI added 2 commits 7 years ago
added 2 commits

3b93cc90 - Use tensorflow global step variable

c0cb1273 - learning rate should go to optimizer

Compare with previous version
Amir MOHAMMADI unmarked as a Work In Progress 7 years ago

unmarked as a Work In Progress
Amir MOHAMMADI @amohammadi started a thread on the diff 7 years ago

doc/user_guide.rst

     >>>
     >>> loss = BaseLoss(tf.nn.sparse_softmax_cross_entropy_with_logits, tf.reduce_mean)
     >>>
     >>> optimizer = tf.train.GradientDescentOptimizer(0.001)
     >>>
     >>> learning_rate = constant(base_learning_rate=0.001)
     >>>
     >>> optimizer = tf.train.GradientDescentOptimizer(learning_rate)

Amir MOHAMMADI assigned to @tiago.pereira 7 years ago

assigned to @tiago.pereira

Tiago de Freitas Pereira @tiago.pereira 7 years ago

Owner

Thanks for the patch @amohammadi, I will merge it right away (I will not wait the builds)

Tiago de Freitas Pereira mentioned in commit 7 years ago

mentioned in commit 65e93dc0

Tiago de Freitas Pereira merged 7 years ago

merged

Please register or sign in to reply