Models: Error running lm_1b model on python3.

[napsternxg]

Please let us know which model this issue is about (specify the top-level directory)

Using model lm_1b

When I run with python3 I get the following errors:

$ python lm_1b/lm_1b_eval.py --mode sample --prefix "I love that I" --pbtxt data/graph-2016-09-10.pbtxt --vocab_file data/vocab-2016-09-10.txt  --ckpt 'data/ckpt-*'
I tensorflow/stream_executor/dso_loader.cc:128] successfully opened CUDA library libcublas.so locally
I tensorflow/stream_executor/dso_loader.cc:128] successfully opened CUDA library libcudnn.so locally
I tensorflow/stream_executor/dso_loader.cc:128] successfully opened CUDA library libcufft.so locally
I tensorflow/stream_executor/dso_loader.cc:128] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:128] successfully opened CUDA library libcurand.so locally
Recovering graph.
Traceback (most recent call last):
  File "/home/napsternxg/Codes/tensorflow_models/bazel-bin/lm_1b/lm_1b_eval.runfiles/__main__/lm_1b/lm_1b_eval.py", line 307, in <module>
    tf.app.run()
  File "/home/napsternxg/anaconda3/lib/python3.5/site-packages/tensorflow/python/platform/app.py", line 43, in run
    sys.exit(main(sys.argv[:1] + flags_passthrough))
  File "/home/napsternxg/Codes/tensorflow_models/bazel-bin/lm_1b/lm_1b_eval.runfiles/__main__/lm_1b/lm_1b_eval.py", line 297, in main
    _SampleModel(FLAGS.prefix, vocab)
  File "/home/napsternxg/Codes/tensorflow_models/bazel-bin/lm_1b/lm_1b_eval.runfiles/__main__/lm_1b/lm_1b_eval.py", line 173, in _SampleModel
    sess, t = _LoadModel(FLAGS.pbtxt, FLAGS.ckpt)
  File "/home/napsternxg/Codes/tensorflow_models/bazel-bin/lm_1b/lm_1b_eval.runfiles/__main__/lm_1b/lm_1b_eval.py", line 88, in _LoadModel
    text_format.Merge(s, gd)
  File "/home/napsternxg/anaconda3/lib/python3.5/site-packages/google/protobuf/text_format.py", line 472, in Merge
    text.split('\n'),
TypeError: a bytes-like object is required, not 'str'

I have also tried running the same using the bazel commands as mentioned in the model README, and I get the same error. Then after referring to #583 I tried using python3 but am still getting the same error.

Here are some additional details:

Python version: 3.5.2 |Anaconda custom (64-bit)| (default, Jul  2 2016, 17:53:06) 
[GCC 4.4.7 20120313 (Red Hat 4.4.7-1)]
Tensorflow version: 0.12.1

[girving]

Can you try this will more recent TensorFlow? I'm curious if it's fixed at 1.0 or HEAD.

[aselle]

Automatically closing due to lack of recent activity. Please update the issue when new information becomes available, and we will reopen the issue. Thanks!

[qinyao-he]

I'm using tensorflow 1.0 on python 3.5 and I have the same issues.

$ bazel-bin/lm_1b/lm_1b_eval --mode sample
--prefix "I love that I"
--pbtxt data/graph-2016-09-10.pbtxt
--vocab_file data/vocab-2016-09-10.txt
--ckpt 'data/ckpt-*'
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcublas.so.7.5 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcudnn.so.4 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcufft.so.7.5 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcurand.so.7.5 locally
Recovering graph.
Traceback (most recent call last):
File "/home/hqy/Code/models/bazel-bin/lm_1b/lm_1b_eval.runfiles/__main__/lm_1b/lm_1b_eval.py", line 307, in
tf.app.run()
File "/home/hqy/.local/lib/python3.5/site-packages/tensorflow/python/platform/app.py", line 44, in run
_sys.exit(main(_sys.argv[:1] + flags_passthrough))
File "/home/hqy/Code/models/bazel-bin/lm_1b/lm_1b_eval.runfiles/__main__/lm_1b/lm_1b_eval.py", line 297, in main
_SampleModel(FLAGS.prefix, vocab)
File "/home/hqy/Code/models/bazel-bin/lm_1b/lm_1b_eval.runfiles/__main__/lm_1b/lm_1b_eval.py", line 173, in _SampleModel
sess, t = _LoadModel(FLAGS.pbtxt, FLAGS.ckpt)
File "/home/hqy/Code/models/bazel-bin/lm_1b/lm_1b_eval.runfiles/__main__/lm_1b/lm_1b_eval.py", line 88, in _LoadModel
text_format.Merge(s, gd)
File "/home/hqy/.local/lib/python3.5/site-packages/google/protobuf/text_format.py", line 472, in Merge
text.split('n'),
TypeError: a bytes-like object is required, not 'str'

please consider reopen this issue @aselle

[maartensap]

I ran into the exact same error. I patched it up adding a simple type check in text_format.py, then ran into some python2to3 issues:
lm_1b_eval.py uses xrange, should be range for python3

[edouardfouche]

I had the same error. I am working with Python 3.5 and Tensorflow '0.12.1'
I replaced s = f.read() at line 86 in lm_1b_eval.py by s = f.read().decode()
Then I replaced the xrange occurences at line 180 and 272 in lm_1b_eval.py by range
Now it works like a charm !

[girving]

Would you be up for submitting a pull request?

[edouardfouche]

Sure, this is done :)

[chencjiajy]

Ｉhad the same error, I run the example of object detection example with Python 3.5 and Tensorflow '1.1'

python create_pascal_tf_record.py --data_dir=/home/user/object_detection_test/VOCdevkit --year=VOC2012 --set=train --output_path=pascal_train.record
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcublas.so.8.0 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcudnn.so.5 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcufft.so.8.0 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcurand.so.8.0 locally
Traceback (most recent call last):
File "create_pascal_tf_record.py", line 183, in
tf.app.run()
File "/root/anaconda3/lib/python3.5/site-packages/tensorflow/python/platform/app.py", line 44, in run
_sys.exit(main(_sys.argv[:1] + flags_passthrough))
File "create_pascal_tf_record.py", line 158, in main
label_map_dict = label_map_util.get_label_map_dict(FLAGS.label_map_path)
File "/home/user/models/object_detection/utils/label_map_util.py", line 122, in get_label_map_dict
label_map = load_labelmap(label_map_path)
File "/home/user/models/object_detection/utils/label_map_util.py", line 107, in load_labelmap
text_format.Merge(label_map_string, label_map)
File "/root/anaconda3/lib/python3.5/site-packages/google/protobuf/text_format.py", line 472, in Merge
text.split('n'),
TypeError: a bytes-like object is required, not 'str'

[edouardfouche]

Does it work if you change label_map_string = fid.read() at line 104 in label_map_util.py to label_map_string = fid.read().decode() ?

[chencjiajy]

It works! thank you!

[hangzh-msft]

adding .decode() is the trick. Thanks @edouardfouche

[hangzh-msft]

Adding .decode() works if you are using TF 1.0. No need for this if you are running TF 1.2.x.

[iosnewbie2016]

I tried to run the object model detection and I run into the same issue. tried changing label_map_string = fid.read().decode(), but still no luck. Note: The line number is not 104 for me, but I made the change in "load_labelmap(path)" method in label_map_util.py

Am using python 3.6, Tensor flow - version 1.0.0, and running on OSX. am running via Anaconda 1.6.3.

[Roc-J]

so anyone could salve the problem?

[Daxel90]

i solve with label_map_string = bytearray(fid.read(), 'utf-8') on line 132