During word2vec trainig, save_inference_model report NoneType error

[lidanqing-intel]

Hi, @luotao1
During training word2vec model, save_inference_model report NoneType error
Build Paddle with following options

cmake ..  -DCMAKE_BUILD_TYPE=Release -DWITH_PYTHON=ON -DPY_VERSION=3.6 -DWITH_AVX=ON -DWITH_DISTRIBUTE=ON

Reproduction the training:

cd Fleet/2.0benchmark/ps/static/word2vec
python -u ../train.py -c benchmark.yaml

The output has following error

Epoch 0 Var LOSS        mean_0.tmp_0      - place: CPUPlace
  - shape: [1]
  - layout: NCHW
  - dtype: float
  - data: [2.74647]
2021-02-24 01:17:59,517 - INFO - Epoch: 0, using time 28.5967116355896 second, ips 36027.394097921366 word/sec.
Traceback (most recent call last):
  File "../train.py", line 245, in <module>
    benchmark_main.run()
  File "../train.py", line 65, in run
    self.run_worker()
  File "../train.py", line 125, in run_worker
    self.infer_target_var)
  File "/home/li/miniconda3/envs/myenv_python3.6/lib/python3.6/site-packages/paddle/distributed/fleet/base/fleet_base.py", line 546, in save_inference_model
    self._runtime_handle._save_inference_model(
AttributeError: 'NoneType' object has no attribute '_save_inference_model'

[paddle-bot-old]

您好，我们已经收到了您的问题，会安排技术人员尽快解答您的问题，请耐心等待。请您再次检查是否提供了清晰的问题描述、复现代码、环境&版本、报错信息等。同时，您也可以通过查看官网API文档、常见问题、历史Issue、AI社区来寻求解答。祝您生活愉快～

Hi! We've received your issue and please be patient to get responded. We will arrange technicians to answer your questions as soon as possible. Please make sure that you have posted enough message to demo your request. You may also check out the API，FAQ，Github Issue and AI community to get the answer.Have a nice day!

[luotao1]

Please see #30560 (comment)

[paddle-bot-old]

Are you satisfied with the resolution of your issue?

YES
No