tensorflow/core/framework/op_kernel.cc:1273] OP_REQUIRES failed at example_parsing_ops.cc:240 : Invalid argument: Key: masked_lm_weights. Can't parse serialized Example.

google-research / bert

TensorFlow code and pre-trained models for BERT

https://arxiv.org/abs/1810.04805

Apache License 2.0

37.42k stars 9.53k forks source link

tensorflow/core/framework/op_kernel.cc:1273] OP_REQUIRES failed at example_parsing_ops.cc:240 : Invalid argument: Key: masked_lm_weights. Can't parse serialized Example. #752

Open yw411 opened 4 years ago

yw411 commented 4 years ago

Tensorflow:1.12 python3.6

when run run_pretraining.py, i meet the error. And when I set the max_predictions_per_seq=5, no error; but when set the max_predictions_per_seq=10, the error happens.

charlesXu86 commented 4 years ago

me too

Broever101 commented 3 years ago

Make sure max_predictions_per_seq matches with the number you set when running create_pretraining_data.py. I had set them differently and I was getting the same issue. When I matched them, the issue was gone.