Input queue is empty when calling next_batch

[chmille3]

I am using this command to run:

python run_summarization.py --mode=decode --data_path=./finished_files/chunked/test_* --vocab_path=./finished_files/vocab --log_root=./pretrained_model --max_enc_steps=400 --max_dec_steps=120 --coverage=1

But I am getting this warning and nothing happens: WARNING:tensorflow:Bucket input queue is empty when calling next_batch. Bucket queue size: 0, Input queue size: 400

Any help is appreciated, thanks!

[chmille3]

I take out the --max_enc_steps=400 and I get this error:

INFO:tensorflow:Failed to load checkpoint from ./pretrained_model/train. Sleeping for 10 secs...

[liyingjiao02]

I am using this command to run:

python run_summarization.py --mode=decode --data_path=./finished_files/chunked/test_* --vocab_path=./finished_files/vocab --log_root=./pretrained_model --max_enc_steps=400 --max_dec_steps=120 --coverage=1

But I am getting this warning and nothing happens: WARNING:tensorflow:Bucket input queue is empty when calling next_batch. Bucket queue size: 0, Input queue size: 400

Any help is appreciated, thanks!

did you find the cause of this problem, i met this problem too.

[chmille3]

It has been awhile since I have worked on this, but I believe the source of the problem was the version of tensorflow I was using with respect to the pre-trained model. I used the pre-trained model for tensorflow 1.2.1. I got mine to work by used that version of tensorflow installed. I hope this helps.

[TanyaChowdhury]

Was anyone able to figure out why this is happening? Changing the tf version didn't work for me.

[charmerphil]

Have you solved it now? Changing the TF version is useless for me.

[TanyaChowdhury]

Yes, I think it was about getting to the exact TF versions for me.