Bert-as-service: TypeError: 'NoneType' object is not iterable

[HeziLiu]

env
win10
python3.6
bert-serving-client/server 1.9.7
tensorflow 1.14.0
and i got this:
E:\Software\Python3.6.5\Scripts>bert-serving-start -model_dir ./uncased_L-12_H-768_A-12/ -num_worker=4
usage: E:\Software\Python3.6.5\Scriptsbert-serving-start -model_dir ./uncased_L-12_H-768_A-12/ -num_worker=4
ARG VALUE

---

       ckpt_name = bert_model.ckpt
     config_name = bert_config.json
            cors = *
             cpu = False
      device_map = []
   do_lower_case = True

fixed_embed_length = False
fp16 = False
gpu_memory_fraction = 0.5
graph_tmp_dir = None
http_max_connect = 10
http_port = None
mask_cls_sep = False
max_batch_size = 256
max_seq_len = 25
model_dir = ./uncased_L-12_H-768_A-12/
no_special_token = False
num_worker = 4
pooling_layer = [-2]
pooling_strategy = REDUCE_MEAN
port = 5555
port_out = 5556
prefetch_size = 10
priority_batch_size = 16
show_tokens_to_client = False
tuned_model_dir = None
verbose = False
xla = False

I:VENTILATOR:freeze, optimize and export graph, could take a while...
WARNING:tensorflow:From e:\software\python3.6.5lib\site-packagesbert_servingserver\helper.py:184: The name tf.logging.set_verbosity is deprecated. Please use tf.compat.v1.logging.set_verbosity instead.

WARNING:tensorflow:From e:\software\python3.6.5lib\site-packagesbert_servingserver\helper.py:184: The name tf.logging.ERROR is deprecated. Please use tf.compat.v1.logging.ERROR instead.

I:GRAPHOPT:model config: ./uncased_L-12_H-768_A-12/bert_config.json
I:GRAPHOPT:checkpoint: ./uncased_L-12_H-768_A-12/bert_model.ckpt
I:GRAPHOPT:build graph...
E:GRAPHOPT:fail to optimize the graph!
Traceback (most recent call last):
File "e:\software\python3.6.5lib\runpy.py", line 193, in _run_module_as_main
"__main__", mod_spec)
File "e:\software\python3.6.5lib\runpy.py", line 85, in _run_code
exec(code, run_globals)
File "E:\Software\Python3.6.5\Scriptsbert-serving-start.exe__main__.py", line 9, in
File "e:\software\python3.6.5lib\site-packagesbert_servingservercli__init__.py", line 4, in main
with BertServer(get_run_args()) as server:
File "e:\software\python3.6.5lib\site-packagesbert_servingserver__init__.py", line 71, in __init__
self.graph_path, self.bert_config = pool.apply(optimize_graph, (self.args,))
TypeError: 'NoneType' object is not iterable

[aszhanghuali]

I have the same question!Have you solved it?

[sivakumarch]

Facing the same issue when I tried to start the server, here is the stacktrace - looks like issue related to tensorflow 2.0 updates. Please let us know if this will be addressed soon.

I:VENTILATOR:[__i:__i: 67]:freeze, optimize and export graph, could take a while...
E:GRAPHOPT:[gra:opt:151]:fail to optimize the graph!
Traceback (most recent call last):
File "skc/environments/.virtualenvs/bert/lib/python3.6/site-packages/bert_serving/server/graph.py", line 41, in optimize_graph
tf = import_tf(verbose=args.verbose)
File "/environments/.virtualenvs/bert/lib/python3.6/site-packages/bert_serving/server/helper.py", line 186, in import_tf
tf.logging.set_verbosity(tf.logging.DEBUG if verbose else tf.logging.ERROR)
AttributeError: module 'tensorflow' has no attribute 'logging'
Traceback (most recent call last):
File "/environments/.virtualenvs/bert/bin/bert-serving-start", line 8, in
sys.exit(main())
File "/environments/.virtualenvs/bert/lib/python3.6/site-packages/bert_serving/server/cli/__init__.py", line 4, in main
with BertServer(get_run_args()) as server:
File "/home/imagen/skc/environments/.virtualenvs/bert/lib/python3.6/site-packages/bert_serving/server/__init__.py", line 71, in __init__
self.graph_path, self.bert_config = pool.apply(optimize_graph, (self.args,))
TypeError: 'NoneType' object is not iterable
$ bert-serving-start -model_dir /bert/bert-models/cased_L-12_H-768_A-12/ -num_worker=1
/home/imagen/skc/environments/.virtualenvs/bert/lib/python3.6/site-packages/bert_serving/server/helper.py:176: UserWarning: Tensorflow 2.0.0 is not tested! It may or may not work. Feel free to submit an issue at https://github.com/hanxiao/bert-as-service/issues/
'Feel free to submit an issue at https://github.com/hanxiao/bert-as-service/issues/' % tf.__version__)

[sivakumarch]

sorry - downgrading to tensorflow 1.x worked, it was due to 2.0 updates. I tried to upgrade the bert_serving scripts using 2.0 tools, but I wasn't successful as there were some errors.

[HeziLiu]

I have the same question!Have you solved it?

I change to another env with
tensorflow 1.13.1
bert-serving-client/server 1.9.9
python3.7
and it works well😀

[HeziLiu]

sorry - downgrading to tensorflow 1.x worked, it was due to 2.0 updates. I tried to upgrade the bert_serving scripts using 2.0 tools, but I wasn't successful as there were some errors.

but with my env with tensorflow 1.14.0 it seems doesn't work....

[ginward]

downgrading to tensorflow version 1.15 seems to work for me

[nonononanana]

check the path of the downloaded model.

[AsmaZbt]

I have the same question!Have you solved it?

I change to another env with
tensorflow 1.13.1
bert-serving-client/server 1.9.9
python3.7
and it works well😀

thank you so much , finally it work for me !

[sprite0153]

I meet that issue to..but I don't want to downgrading,Is there have any other settlement work for 2.0

[harshitbh]

I am still facing the above issue have tried couple of solutions.
Version - tensorflow-cpu = 1.13.1
Bert_serving_client = 1.9.9
Bert_serving_server = 1.9.9

[harshitbh]

I am running the below command after locating at model_dir
command bert-serving-start -model_dir uncased_L-2_H-128_A-2\ -cpu
files present in model_dir :bert_config.json,bert_model.ckpt.data-00000-of-00001,bert_model.ckpt.index,vocab.txt
have tried with providing absolute path till model directory too
system config - 12 gb ram.
let me know how could I resolve this.Thanks in advance

[Jaswanth-Batturi]

I am getting the same problem in windows, although I tried all the methods.
"TypeError: 'NoneType' object is not iterable".
Please somebody help me.

[wingzzzz]

你们模型目录写错啦

[Doracoco]

你们模型目录写错啦
那应该怎么写呢，没有发现哪里写错了

[wingzzzz]

你们模型目录写错啦
那应该怎么写呢，没有发现哪里写错了

你试试绝对路径

[Doracoco]

你们模型目录写错啦
那应该怎么写呢，没有发现哪里写错了

你试试绝对路径

还是不行的，我的绝对路径还是不行，不过还是谢谢你啦

[nonononanana]

加速 入关

[wingzzzz]

你们模型目录写错啦
那应该怎么写呢，没有发现哪里写错了

你试试绝对路径

还是不行的，我的绝对路径还是不行，不过还是谢谢你啦

有可能是tensorflow版本过高，api不兼容，只支持1.x.x版本的。报错信息跟进去，可以看到源码，大于1的版本会报一个未测试验证的警告。