variable does not exist
djjandXMU opened this issue · 10 comments
When I run the main.py,I find that ValueError: Variable d_model/d_block_0/downconv/W/RMSProp/ does not exists.
Could you help me?
Best regards
@djjandXMU I am also getting the same error. I have tried https://stackoverflow.com/questions/43183850/tensorflow-valueerror-variable-does-not-exist-or-was-not-created-with-tf-get-v but nothing seems to work.
Did you get the solution??
Best Regards
The given error can be solved by changing a fraction of a code...
In main.py just change the optimizer function from tf.Adamoptimizer() to tf.GradientDescentOptimizer().
but the tradeoff to change main.py is traning time.
model.py line144 self.d_opt = d_opt.apply_gradients(avg_d_grads)
model.py line145 self.g_opt = g_opt.apply_gradients(avg_g_grads)
Maybe you can replace line144-145 with the following code.
with tf.variable_scope(tf.get_variable_scope(), reuse=tf.AUTO_REUSE):
self.d_opt = d_opt.apply_gradients(avg_d_grads)
self.g_opt = g_opt.apply_gradients(avg_g_grads)
@AnilMandliya @djjandXMU i have fixed this issue in my fork: https://github.com/fcakyon/segan
@fcakyon
Sorry to bother you, I used your fork to train this model on a no gpu machine，When I run train_segan.sh，I met the following error:
tensorflow.python.framework.errors_impl.NotFoundError: No registered 'TemporaryVariable' OpKernel for XLA_CPU devices compatible with node {{node device_1/gradients/AddN_14/tmp_var}}
Could you please help me figure out what problem happens? Any reply will be appreciated.
@woshiZS unfortunately I no longer work on this project, I hope someone else can help you 👍
@AnilMandliya @djjandXMU我已经在我的 fork 中解决了这个问题：https://github.com/fcakyon/segan
你好，我发现在你的readme文件中提供的预训练模型链接点不开，好像失效了，能重新发我一份吗，麻烦了
@AnilMandliya @djjandXMU我已经在我的 fork 中解决了这个问题：https://github.com/fcakyon/segan
你好，我发现在你的readme文件中提供的预训练模型链接点不开，好像失效了，能重新发我一份吗，麻烦了