LoginSignup
0
0

More than 3 years have passed since last update.

How to see the training logs

Last updated at Posted at 2020-04-24

environment

  • Macos Catalina 10.15.1
  • python 3.6.8
  • pip 20.0.2

install tensorboard

# use list command to check if you have already installed tensorboard
# if there is no tensorboard, use pip to install
# you can check your tensorflow version firstly and install the corresponding version of tensorboard

$ pip list
$ pip search tensorboard
tensorboard (2.2.1)                             - TensorBoard lets you watch
                                                  Tensors Flow
  INSTALLED: 1.12.2
  LATEST:    2.2.1
neptune-tensorboard (0.3.9)                     - Neptune Tensorboard
jupyter-tensorboard (0.2.0)                     - Start tensorboard in
                                                  Jupyter! Jupyter notebook
                                                  integration for tensorboard.
tensorboard-chainer (0.4.0)                     - Log TensorBoard events with
                                                  chainer
tensorboard-pytorch (0.7.1)                     - Log TensorBoard events with
                                                  pytorch
convolut-tensorboard (0.0.2)                    - Tensorboard integrations for
                                                  convolut framework
tensorboard-easy (0.2.3)                        - A tensorflow-independent
...
$ pip install tensorboard==1.12.2

prepare the training logs

# at first you may choose to create and save the training log
# and the saving path may not include character path

tensorboard --logdir the folder path of saving logs files
$ tensorboard --logdir 
/Users/banshijou/.pyenv/versions/3.6.7/envs/tensorflow1.12/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:523: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint8 = np.dtype([("qint8", np.int8, 1)])
/Users/banshijou/.pyenv/versions/3.6.7/envs/tensorflow1.12/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:524: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint8 = np.dtype([("quint8", np.uint8, 1)])
/Users/banshijou/.pyenv/versions/3.6.7/envs/tensorflow1.12/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint16 = np.dtype([("qint16", np.int16, 1)])
/Users/.pyenv/versions/3.6.7/envs/tensorflow1.12/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:526: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint16 = np.dtype([("quint16", np.uint16, 1)])
/Users/.pyenv/versions/3.6.7/envs/tensorflow1.12/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:527: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint32 = np.dtype([("qint32", np.int32, 1)])
/Users/.pyenv/versions/3.6.7/envs/tensorflow1.12/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:532: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  np_resource = np.dtype([("resource", np.ubyte, 1)])
TensorBoard 1.12.2 at http://PC_name:6006 (Press CTRL+C to quit)

open the tensorboard at browser

close the tensorboard

press CTRL + C to release the port 6006

/Users/.pyenv/versions/3.6.7/envs/tensorflow1.12/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:532: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  np_resource = np.dtype([("resource", np.ubyte, 1)])
TensorBoard 1.12.2 at http://PC_name:6006 (Press CTRL+C to quit)
^C(tensorflow1.12) PC_name:~ user$ 

you can also check the tensorflow HP for more message

0
0
0

Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up
0
0