did anybody tried to load the pretrained model provided? about refresh HOT 7 CLOSED

i tried this code
with tf.Session() as sess:
model = MY_Model(sess, len(vocab_embed_object.vocab_dict)-2)
selected_modelpath="./pretrained/model.ckpt.epoch-11"
print("Reading model parameters from %s" % selected_modelpath)
model.saver.restore(sess, selected_modelpath)
print("Model loaded.")

got this error

Unable to open table file ./pretrained/model.ckpt.epoch-11: Data loss: block checksum mismatch: perhaps your file is in a different file format and you need to use a different restore operator?

using python 2 and tensorflow 0.10.0rc0

from refresh.

You could use the author's code for loading the model. It works for me with python2 and tensorflow 0.10.0

from refresh.

from refresh.

I did used the author's code and only changed the directory name for the pre-trained model folder etc.

Are you running this on Windows? It seems that weird errors could occur with tensorflow on Windows.
Can you try to create a virtual environment then install the dependencies.

from refresh.

Can you try tensorflow 0.12 instead of 0.10.0rc0?

Check this link also: https://stackoverflow.com/questions/41048819/how-to-restore-a-model-by-filename-in-tensorflow-r12

from refresh.

turns out that the model file was not uploaded correctly on the instance,worked fine after reuploading .Thanks for the help

from refresh.

Thanks for helping one another guys. I have been a bit busy with the move.

from refresh.