Tensorboard

The 2.7 minor series tracks TensorFlow 2.7.

Features

- Time Series plugin
- Run selection now is based on regex filter (5252)
- Run match logic matches run name and alias (5334, 5351)
- Prepare Time Series for promotion to the first tab (5291)
- Improve/persist tag filter in the URL (5249, 5236, 5263, 5265, 5271, 5300)
- Show sample count on image cards (5250)
- Keep all digits for step values (5325)
- Remove pinned view while filtering (5324)
- Show relative time in tooltip (5319)
- UI: style improvements, adjust scroll position
- Core
- Resizable run table sidebar (5219)
- Support for fsspec filesystems (5248)
- Hparams
- Treat no data as an empty experiment rather than an error (5273)
- Add tf.stop_gradient in tf.summary.histogram (5311) - thanks [allenlavoie](https://github.com/allenlavoie)

Bug fixes

- Darkmode improvements and fixes (5318)
- Time Series
- Improve visibility logics (5234, 5235)
- Reset PluginType filter when selected all (5272)
- PR curve plugin: display correct thresholds (5191)
- Line chart
- Recreate charts upon fatal renderer errors (5237)
- Fix zoom interaction (5215)
- Skip axis label render based on visibility (5317)
- Dropdown ui fixes (5194, 5199, 5242)
- Navigation handling (5223, 5216)
- Documentation
- document the Time Series dashboard (5193)
- Update README.md to include no-data example (5163)

The 2.6 minor series tracks TensorFlow 2.6.

Features

- Added dark mode
- Some user settings are now kept in local storage and are persisted after page reload, including:
- Time Series `scalars smoothing`, `tooltip sorting method`, `ignore outliers in chart scaling`
- Settings dialog `reload data`, `reload period`, `pagination limit`
- Dark mode
- Time Series
- Improved positioning of the `fit` button in scalar chart (4856)
- Improved selection of runs when new runs arrive (4888)

Bug fixes

- Fixed bug where some plugins were not appearing in plugins list (4849)
- Fixed subtle bugs in navigation (4974)
- Stopped storing default values for settings on the URL (5030)
- Graphs
- Fixed parsing of `_output_shapes` attr (4867)
- HParams
- Fixed parallel coordinate layout (4988)
- Projector
- Fix KNN algorithm and, by extension, T-SNE and UMAP embeddings (5063)
- Make opaque sprites opaque (4921) - thanks [tyhenry](https://github.com/tyhenry)
- Make transparent sprites transparent (5149) - thanks [canbakiskan](https://github.com/canbakiskan)
- Scalars
- Improved rendering of small major axis numbers (5010)
- Fixed axis label in Firefox (5078)
- Time Series
- Fixed rendering of filter text areas (4938)
- Improved rendering of small major axis numbers (5010)
- Improved tooltip rendering (5003)
- Fixed scrolling of runs selector (5020)
- Make runs selector header sticky on scroll (5024)
- Improved rendering of step axis to use SI units (5015)
- Documentation
- Fixed and cleaned documentation and demos during TensorBoard team fixit.
- Fixed debugger v2 documentation (4843) - thanks [kevint324](https://github.com/kevint324)

TensorBoard.dev updates

- Improve upload throughput for scalar summaries (4825)
- Fixed bug where experiment name and description were not displayed in `tensorboard dev list` (4912)
- Fixed bug where reading from remote directories did not work in `--one_shot` mode (4909)

Breaking changes

- TimeSeries plugin no longer supports `?fastChart=false` fallback to old chart renderer.

Features

- New data loading mode: typically loads between 100× and 350× faster
- On by default when applicable; pass `--load_fast false` to disable
- For details, or to provide feedback, see 4784
- Time Series
- Improved line charts with GPU acceleration (4695)
- On by default; use `?fastChart=false` to fallback to the older version. Please give us feedback before you use the fallback mechanism.
- Support manual extent changes (4711)
- Monotonic option for "zigzag" charts (4696)
- Graphs
- Added toggle to disable high-degree node extraction (4722)
- "Color by structure" is now opt-in for large graphs; improves load performance by up to 77% in some cases (4742)
- Combine graph by prefixing with unique name (4334)
- Text
- Support fenced code blocks in Markdown (4585)
- Histograms
- Most recent step gets the most salient color. (4374)
- Support http server port reuse (4616) – thanks zuston.

Bug fixes

- Graphs
- Fixed "Download as PNG" button (4759)
- Fixed "Ungroup this series" button (4817)
- Now collapses more series (abc1, abc2, …, abc5 → abc[1-5]) (4803)
- Core
- Removed scrollbars on no data views for plugins (4525)
- Error message is JS is disabled (4401)
- Projector: fix the KNN algorithm caused projections to be incorrect. (4687)
- Windows: fixed the issue where TensorBoard notebook magic would hang when first launched (4407, 4300)
- Time series: Time axis on scalar chart read walltime incorrectly (4541)
- Graph events inconsistently evicted after session log START event (4743)

TensorBoard.dev updates

- Text summaries now uploaded

Breaking changes

- Projector plugin will not work with `--logdir_spec` (4494) (may not have worked before)
- Drop support for negative `--reload_interval`, which instructed TensorBoard to not read any data at all

Bug fixes

- Fixed `--path_prefix` handling (4423)
- Removed `frame-ancestors *` CSP directive for compatibility with Electron
embeds (4332) - thanks [joyceerhl](https://github.com/joyceerhl)

The 2.4 minor series tracks TensorFlow 2.4.

Features

- Improved performance for scalar charts with many runs
- Up to 50% faster network fetch times in some cases (4050)
- Up to 90% faster paint time in some cases (4053)
- 🧪 **Experimental** Time Series dashboard
- View scalars, histograms, and images side-by-side in a combined view
- Customize the color of specific runs
- Pin specific charts/images/histograms and share a custom view of your data
- No additional logging required

TensorBoard.dev updates

- Added support for uploading Hparams (3916)
- Try `tensorboard dev upload` on a logdir containing
[hparams][hparam-tutorial]!

[hparam-tutorial]: https://www.tensorflow.org/tensorboard/hyperparameter_tuning_with_hparams

Bug fixes

- Docs: fixed image summary tutorial (4206)
- Projector Plugin: fixed bookmark loading (4159), thanks aknoerig!
- Graphs: updated TPU-compatible ops list (4024)

Deprecations

TensorBoard features that depend on TensorFlow APIs now require TensorFlow 2.x
installed. Running TensorBoard 2.4.0+ with TensorFlow 1.x installed will not be
supported.

Please note that this does not affect data already written to disk; summaries
emitted by the TF 1.x tf.summary API are still readable and fully supported by
the latest versions of TensorBoard.

Support for Python 3.5 is dropped.

Plugin dashboards have been removed:
- Beholder: see 3843
- Debugger V1: replaced by [Debugger V2][debugger_v2_tutorial]

[debugger_v2_tutorial]: https://www.tensorflow.org/tensorboard/debugger_v2

The 2.3 minor series tracks TensorFlow 2.3.

Features

- The 30 sec default reload period is now customizable in the Settings UI
(2794)
- 🧪 **Experimental** Debugger V2 is now available; see the
[tutorial][debugger-v2-tutorial] on how to use the experimental TensorFlow
APIs to spot NaN's in tensors, view graphs generated from executions, and the
related lines in the Python source code (3821)

TensorBoard.dev updates

- Added support for showing the Distributions tab (3762)
- Uploader now displays data statistics in the console while uploading data
(3678)
- Added new uploader command line flags (3707)
- `--dry_run`: causes the uploader to only read the logdir and display
statistics (if `--verbose` is the default 1) without uploading any data to
the server
- `--one_shot`: causes the uploader to exit immediately after all existing
data in the logdir are uploaded; this mode prints a warning message if the
logdir doesn't contain any uploadable data
- Upload button in the header offers a convenient, copyable command
- 🧪 **Experimental** DataFrame API: You can now read Scalars data from
TensorBoard.dev as a Pandas DataFrame (learn more [here][dataframe-tutorial])

[debugger-v2-tutorial]: https://www.tensorflow.org/tensorboard/debugger_v2
[dataframe-tutorial]: https://www.tensorflow.org/tensorboard/dataframe_api

Bug fixes

- Projector plugin
- Shows data when logs exist in both logdir root and subdirectory (3694)
- Fixed incorrect embeddings from TF2 checkpoints (3679)
- Added support for binary format, with 2x speedup loading large tensors in
some cases (3685) - thanks [RustingSword](https://github.com/RustingSword)
- Added [Colab tutorial][projector-colab] for Projector plugin (3423)
- Notebooks
- Increased port scanning from 10 to 100 to better support multi-tenant
Notebooks (3780) - thanks [jerrylian-db](https://github.com/jerrylian-db)
- Add proxy (e.g. jupyter-server-proxy) support for %tensorboard magics
(3674) - thanks [zac-hopkinson](https://github.com/zac-hopkinson)
- Set the TENSORBOARD_PROXY_URL environment variable
`export TENSORBOARD_PROXY_URL="/proxy/%PORT%/"`
- Dynamic plugins (Projector, Fairness Indicators, Profiler, What-If Tool)
appear when TensorBoard is launched programmatically via Python (3695)
- Fixed download links in Custom Scalars (3794)
- Updated broken docs (3440, 3459, 3561, 3681) - thanks
[LexusH](https://github.com/LexusH),
[ManishAradwad](https://github.com/ManishAradwad),
[ricmatsui](https://github.com/ricmatsui),
[robertlugg](https://github.com/robertlugg)
- Better handling of S3-related InvalidRange errors (3609) - thanks
[ahirner](https://github.com/ahirner)
- Fixed deprecated numpy usage (3768) - thanks
[lgeiger](https://github.com/lgeiger)

[projector-colab]: https://www.tensorflow.org/tensorboard/tensorboard_projector_plugin

Deprecations

- Beholder will be removed in a future release (3843)
- Debugger (V1) will be removed in a future release, in favor of the
aforementioned V2 version

Misc

The frontend now uses Angular (replaces the Polymer entry point, which will be
removed in a future release; still visible at the `/legacy.html` endpoint)
(3779). If you observe any bugs that do not reproduce under `/legacy.html`,
please file an issue.

For dynamic plugins, please see their respective pages
([Fairness Indicators][fairness-docs], [Profiler][profiler-docs],
[What-If Tool][wit-docs]).

[fairness-docs]: https://github.com/tensorflow/fairness-indicators/commits/master
[profiler-docs]: https://github.com/tensorflow/profiler/commits/master
[wit-docs]: https://github.com/PAIR-code/what-if-tool/blob/master/RELEASE.md