- Supports tensorshape information in graph visualization. Drop support for 0.3.1
- Adds add_video function
- Supports pytorch 0.3.1 (hacky)
- Supports graph (the pretty one)
- Supports markdown for add_text function
- It's ready to log precision recall curve (needs tensorboard>=0.4)
- Adds context manager for the SummaryWriter class
- Package name renamed to tensorboardX to fix namespace confliction with tensorflow's tensorboard
- Supports multi-scalars and JSON export
- Multiple Embeddings in One Experiment
- Supports Chainer and mxnet
- fixed incorrect image<->label pairing in embedding function (#12)
- unifies API call and adds docstring. Documentation is available at: http://tensorboard-pytorch.readthedocs.io/
- remove tensorflow dependency for embedding function
- fixed incorrect image<->label pairing in embedding function (#12)
- unifies API call and adds docstring. Documentation is available at: http://tensorboard-pytorch.readthedocs.io/
- add travis test (py2.7, py3.6)
- add support for python2 (in PyPI)
- supports embedding
- supports graph summary
- fixed np.histogram issue
- supports text summary
- supports audio summary
- simplifies add_image API
- speed up add_histogram API by 35x
- First commit. Reference:
https://github.com/TeamHG-Memex/tensorboard_logger https://github.com/dmlc/tensorboard