Open Jatapiaro opened 7 years ago
I got the same issue with Python 3.6 with tensor 1.1.0
I found a solution for it. You should check my code https://github.com/Jatapiaro/TwinkerBot_Revival Twitter traning starts the bot training, that's where I got the error. Copy paste my seq2seq_wrapper code, use python 3 and please use tensorflow 1.0.0
then I receive this... thank you
@dexterchan Are you shure that you're using Tensorflow 1.0.0 and python3? You can check your tensorflow version using the command below: python3 -c "import tensorflow as tf; print(tf.version)"
Sorry, forget to upgrade my tensorflow.... it is running training now... thank you for your support.... :)
@dexterchan You're welcome :) And please be careful with this https://github.com/suriyadeepan/practical_seq2seq/issues/2 I recommend you to left your training running as most as you can, I've lost one day trying to restore sessions but it doesn't work, for some reason begins again.
To anyone trying to run python3 -c "import tensorflow as tf; print(tf.version)"
, you'll get a AttributeError
, version
has been renamed to VERSION
, so, now you've to run python3 -c "import tensorflow as tf; print(tf.VERSION)"
Using Tensorflow 1.0.0 fix this issue.
To downgrade the gpu version pip3 install tensorflow-gpu==1.0.0
To downgrade the cpu version pip3 install tensorflow==1.0.0
@anthdm Thank you ! That works for me! Downgrade the Tensorflow to 1.0.0.
Is there any other way ?I wouldn't downgrade the tensorflow to 1.0.0
I'm running the twitter chatbot code but I'm getting the next error: "can't pickle _thread.lock objects"
What does that means?