Emmental

Latest version: v0.1.1

Safety actively analyzes 683530 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

Page 2 of 2

0.0.5

-------------------

Added
^^^^^

* `senwu`_: Add `checkpoint_all` to controll whether to save all checkpoints.
* `senwu`_: Support `CosineAnnealingLR`, `CyclicLR`, `OneCycleLR`, `ReduceLROnPlateau`
lr scheduler.
* `senwu`_: Support more unit tests.
* `senwu`_: Support all pytorch optimizers.
* `senwu`_: Support accuracyk metric.
* `senwu`_: Support cosine annealing lr scheduler.

Fixed
^^^^^

* `senwu`_: Fix multiple checkpoint_metric issue.

0.0.4

-------------------

Added
^^^^^

* `senwu`_: Log metric dict into log file every trigger evaluation time or full epoch.
* `senwu`_: Add `get_num_batches` to calculate the total number batches from all
dataloaders.
* `senwu`_: Add `n_batches` in `EmmentalDataLoader` and `fillup` in `Scheduler` to
support customize dataloader.
* `senwu`_: Add overall and task specific loss during evaluating as default.
to support user needs for clear checkpoins.
* `senwu`_: Add `min_len` and `max_len` in `Meta.config` to support setting sequence
length.
* `senwu`_: Add overall and task specific loss during evaluating as default.
* `senwu`_: Calculate overall and task specific metrics and loss during training.
* `senwu`_: Add more util functions, e.g., array_to_numpy, construct_identifier,
and random_string.
* `senwu`_: Enforce dataset has uids attribute.
* `senwu`_: Add micro/macro metric options which have split-wise micro/macro average
and global-wise micro/macro average. The name for the metrics are:

::

split-wise micro average: `model/all/{split}/micro_average`
split-wise macro average: `model/all/{split}/macro_average`
global-wise micro average: `model/all/all/micro_average`
global-wise macro average: `model/all/all/macro_average`

*Note*: `micro` means average all metrics from all tasks. `macro` means average all
average metric from all tasks.

* `senwu`_: Add contrib folder to support unofficial usages.

Fixed
^^^^^

* `senwu`_: Correct lr update for epoch-wised scheduler.
* `senwu`_: Add type for class.
* `senwu`_: Add warning for one class in ROC AUC metric.
* `senwu`_: Fix missing support for StepLR and MultiStepLR lr scheduler.
* `senwu`_: Fix missing pytest.ini and fix test cannot remove temp dir issue.
* `senwu`_: Fix default train loss metric from `model/train/all/loss` to
`model/all/train/loss` to follow the format `TASK_NAME/DATA_NAME/SPLIT/METRIC`
pattern.

Changed
^^^^^^^

* `senwu`_: Change default grad clip to None.
* `senwu`_: Update seed and grad_clip to nullable.
* `senwu`_: Change default class index to 0-index.
* `senwu`_: Change default ignore_index to None.
* `senwu`_: Change the default counter unit to epoch.
* `senwu`_: Update the metric to return one metric value by default.

Removed
^^^^^^^

* `senwu`_: Remove `checkpoint_clear` argument.

.. _Unreleased: https://github.com/senwu/emmental/compare/v0.1.1...main
.. _0.0.4: https://github.com/senwu/emmental/compare/v0.0.3...v0.0.4
.. _0.0.5: https://github.com/senwu/emmental/compare/v0.0.4...v0.0.5
.. _0.0.6: https://github.com/senwu/emmental/compare/v0.0.5...v0.0.6
.. _0.0.7: https://github.com/senwu/emmental/compare/v0.0.6...v0.0.7
.. _0.0.8: https://github.com/senwu/emmental/compare/v0.0.7...v0.0.8
.. _0.0.9: https://github.com/senwu/emmental/compare/v0.0.8...v0.0.9
.. _0.1.0: https://github.com/senwu/emmental/compare/v0.0.9...v0.1.0
.. _0.1.1: https://github.com/senwu/emmental/compare/v0.1.0...v0.1.1
..
For convenience, all username links for contributors can be listed here

.. _senwu: https://github.com/senwu
.. _lorr1: https://github.com/lorr1

Page 2 of 2

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.