Martins.Abilevs
Martins.Abilevs
but is there solution ?? because i want to inject in training new conversations and voc will grow anyway ... same to tensor size my plan is to incarase from...
not only crash ... ya lost previous training data :D
and next i want to split by sylabies all words .... in far future when i get some nvidia gfx card with ton of cudcores
model.py !! be avare seq2seq in3.5 is moved to : tf.contrib.legacy_seq2seq.embedding_rnn_seq2seq and loss to: tf.contrib.legacy_seq2seq.sequence_loss
there no such constant amount of iterations. forget it !!! it is finished when you get expected result in testing. training will run until you hit ctrl+c
once i got the same warning and crash ... i can`t remmember what i did exactly ... but you need to get rid of this warning !!!
https://github.com/MartinAbilev/tenzorflow-chat-testdrive it is one of multiple versions i got. it is working for me at this moment it is on python3
same struggle python 3 already lot of time wasted to wait parsing and in final get error about binary or string file types ... please help to get this mess...
there no any binary file in this example at all !!! all is text files. !!! cornel converations is text, output vocabs is text. now you reading text file as...
looks like not enough training.