提交

提交图

1485 次代码提交

作者 SHA1 备注 提交日期
Jonas Mueller a24dda5da8 clarify features for multiannotator tutorial 2022-09-16 11:22:02 -07:00
Jonas Mueller 7256cd3546 More improvements to token classification code and documentation (#452)
* improved token docs/code

* format paragraphs in docstrings

* fix typo

Co-authored-by: Elías Snorrason <eliassno@gmail.com>
2022-09-16 14:04:45 +00:00
Jonas Mueller 688c39c202 hide redundant display_example in audio tutorial 2022-09-16 04:09:15 -07:00
Jonas Mueller 5d900200e5 mention make_data in multiannotator tutorial 2022-09-16 03:46:03 -07:00
Jonas Mueller 2959210c95 more links in keras wrapper docstrings 2022-09-15 20:20:31 -07:00
Jonas Mueller 0657b14792 add links in keras wrapper docs 2022-09-15 19:20:37 -07:00
Elías Snorrason bc2dabb764 Change output of display_issues (#450) 2022-09-15 18:05:46 -07:00
Elías Snorrason 95b0742342 Make softmin_sentence_score a private function (#449)
This scoring function is only used internally. Might as well be private.
2022-09-15 23:19:15 +00:00
Elías Snorrason cdf6a147af ignore links to pradyunsg when validating HTML (#446) 2022-09-15 15:21:12 -07:00
Curtis G. Northcutt 1b813d9666 fix bug in hard-coded test. generalize the test (#448)
* fix bug in hard-coded test. generalize the test

* 🐛 cast rounded num_issues to int

np.rint outputs an array of the same shape and type as its input. num_label_issues is expected to return an integer.

Co-authored-by: Elías Snorrason <eliassno@gmail.com>
2022-09-15 18:16:56 -04:00
Jonas Mueller 1b76f4ff1b Improve docstrings for keras wrapper (#447)
* improved docstrings for keras wrapper

* hide inheritance from keras wrapper docstrings

* black formatting
2022-09-15 15:36:35 -04:00
Curtis G. Northcutt 3ff8fab2d9 make num_label_issues = cj calibrated offdiag sum (#445)
The discrepancy occurs because cj calibrated maintains a guarantee that it will perfectly count every example in the dataset (assumes no out of distribution examples).

You can see this when by following this change of methods:

* see `calibrate_confident_joint`, specifically this line here: https://github.com/cleanlab/cleanlab/blob/master/cleanlab/count.py#L172
* round here https://github.com/cleanlab/cleanlab/blob/9240a8c56d020abec8566a0a56d22dcb99d32528/cleanlab/internal/util.py#L209
* the actual rounding occurs here https://github.com/cleanlab/cleanlab/blob/9240a8c56d020abec8566a0a56d22dcb99d32528/cleanlab/internal/util.py#L174

the fix to make `num_label_issues` the same is just to round instead of flooring.

see the two GREEN columns in the attached image

> [nit] consider `np.rint()` instead of `np.round()` for clarity

good call. updated
2022-09-15 11:44:27 -07:00
Jonas Mueller 9240a8c56d rst file for keras wrapper 2022-09-15 10:08:12 -07:00
Elías Snorrason 5167d23b71 Update links to example notebooks (#444)
Related to cleanlab/examples#21
2022-09-15 11:10:33 -04:00
Jonas Mueller a4ad6460a2 Format return docstrings and add typing (#437)
* typing and return docstrings

* address typing complaints

* allow redefinition

* typing complains for pandas

* .values -> to_numpy()
2022-09-15 11:46:14 +00:00
Jonas Mueller 7ccdb5491f Add keras wrapper to docs (#443)
* Delete huggingface_keras_classifier.py

* Update keras.py

* readme language, remove huggingfacekerasclassifier

* language improvements

* reorder modules

* make keras.py appear in docs
2022-09-15 00:47:31 -07:00
Ulyana 61f650e357 Merge pull request #426 from ulya-tkch/ulya-outlier-array-like
Updated labels to allow array_like
2022-09-14 21:21:33 -07:00
Elías Snorrason fd50f6b9c6 Deploy docs for token classification module (#438)
* docs: 📝 add usage example for get_sentence

Also remove types in docstrings.

* docs: 📝 add usage examples for filter_sentence

* docs: 📝 add page for internal.token_classification_utils

* docs: 📝 remove name of return variable

auto-tagging won't work correctly for the return value if it's not a tuple

* 📝 add pages for modules in cleanlab.token_classification

* fix indentation

* 📝 add token classification api to toctree

* 🎨 format docstrings

* 📝 add usage examples

* update example comment

* 📝 add usage examples

* 📝 add return variable name in docstring

* 📝 remove some hard-coded params types in rank module

Keep more complex parameters hard-coded for now. E.g. "enum"-like parameters.

* 🎨 remove indentation of parameters in docstrings

* 🎨 fix indentation

* colored docstring typo + formatting

* typo: occurence

* probs_merged docstring typo + formatting

Co-authored-by: Jonas Mueller <1390638+jwmueller@users.noreply.github.com>
2022-09-14 14:59:26 -07:00
Hui Wen d74dda0ed9 fix code and md inconsistency (#442) 2022-09-14 14:45:09 -07:00
Elías Snorrason cb5bfe8081 Add instructions for using sphinx-build (#441)
* 📝 add instructions for using sphinx-build

This is useful for building docs for un-committed changes.

* more clarification on uncommitted sphinx build

* another clarification that we dont have to commit

Co-authored-by: Jonas Mueller <1390638+jwmueller@users.noreply.github.com>
2022-09-14 13:23:47 -07:00
Hui Wen 5603b1c8a6 Extend KerasWrapper to Functional API (#434)
* add keras functional api wrapper

* add breif docstring

* minor unittest edit

* edit unittest

* update docstring

* make unittest less strict
2022-09-14 13:10:12 -07:00
Elías Snorrason 968310af37 Strip input prompts when copying code snippets (#439)
Configuration provided in https://sphinx-copybutton.readthedocs.io/en/latest/use.html#using-regexp-prompt-identifiers
2022-09-14 14:06:04 +00:00
Jonas Mueller 9b78b703d5 black formatting 2022-09-14 00:00:18 -07:00
Jonas Mueller bf000d551b update labels typing to be LabelLike 2022-09-13 23:55:35 -07:00
Hui Wen 7ab4cbbf23 minor docstring edits (#436) 2022-09-13 19:01:00 -07:00
Elías Snorrason 729c9f3068 Add autodoc-typehints extension for sphinx (#412)
* add autodoc-typehints extension for sphinx

With the extension: Function parameters are auto-tagged in docs, based on the function signatures.

Resolves #398

* disable use of :rtype: role by napoleon

output return type inline with description (https://sphinxcontrib-napoleon.readthedocs.io/en/latest/sphinxcontrib.napoleon.html#sphinxcontrib.napoleon.Config.napoleon_use_rtype)
2022-09-13 13:12:28 -07:00
Jonas Mueller 84af558e85 hide more code in image tutorial, organize imports 2022-09-13 12:39:43 -07:00
Jonas Mueller a9e3c0e5d2 emphasize features must be numeric 2022-09-13 09:37:21 -07:00
Jonas Mueller 6a0e3bc461 Improvements to token_classification tutorial (#419)
* edits to token tutorial

* fix typos

* given_labels -> labels in quickstart block

* add IOB2-formatting clarification

Co-authored-by: Elías Snorrason <eliassno@gmail.com>
2022-09-13 09:34:33 -07:00
Jonas Mueller b9ab4ba57b suppress tensorflow warning logs in tutorials if not properly installed (#432)
* remove ineffective TF suppression lines

* suppress tensorflow logs in workflows tutorial

* formatting

* hidden cell comment

* move suppress code to hidden cell

* move suppression code to first hidden cell

* move suppression code to first cell

* supress tf logs in tabular tutorial

* optional code to hidden cells in audio tutorial

* suppress tf logs in multiannotator tutorial

* suppress tf logs in token tutorial

* move cross-val tutorial to appear last

* move cross-val tutorial to be last in main index

* move outlier before multiannotator in API list

* remove newline at end of cell
2022-09-13 00:47:21 -07:00
Jonas Mueller 15899f8f20 try to suppress tensorflow logs pt2 (#431)
* move env variable setting before tf import

* suppress tf logs in faq
2022-09-12 18:03:00 -07:00
Ulyana 13d6cfc9ce Used internal function for typechecking labels 2022-09-12 16:54:07 -07:00
Jonas Mueller a0276ca587 Text tutorial improvements (#429)
* more tf supression

* formatting

* move imports to top
2022-09-11 14:46:04 -07:00
Elías Snorrason 9a109182b1 Update argument given_labels -> labels in token classification module (#423)
* given_labels -> labels

only applies to the argument to `display_issues`

Fixes #418

* update kwargs in test for display_issues

* apply black formatter
2022-09-11 12:38:38 -07:00
Jonas Mueller f7102db902 Add links to examples notebooks where appropriate (#428)
* links to OOD examples notebook

* iterative notebook title in its link
2022-09-11 12:24:51 -07:00
Jonas Mueller c307e9714f update workflows for post 2.0 (#427)
* update workflows for post 2.0

* formatting

* link for cleanlearning
2022-09-11 00:27:57 -07:00
Ulyana a6fb96f082 Add test 2022-09-09 16:58:42 -07:00
Ulyana 5243b48b00 Updated labels to array_like (list is ok) 2022-09-09 16:21:29 -07:00
Ulyana 43a795cb0c Update DEVELOPMENT.md with Relative Linking (#425)
* Update DEVELOPMENT.md with Relative Linking

Added instructions for relative linking between docs and tutorials

* more link instructions

Co-authored-by: Jonas Mueller <1390638+jwmueller@users.noreply.github.com>
2022-09-09 15:43:11 -07:00
Jonas Mueller d4deced278 add fork instruction 2022-09-09 13:50:33 -07:00
Ulyana 31d2618560 Improve outliers docstring notations (#424)
* Docstring remove args test

* exclude members test

Please enter the commit message for your changes. Lines starting

* Adding params as class instance attribute

* Added autoattribute: params

* Adding autoattribute under class

* Removed no-exclude-members and addded params to members

* autoattribute with indent

* Added annotation params

* Revert to removing default params only

* Removed params as instance var listing

* Consist italics vs code syntax in docstring
2022-09-09 12:55:29 -07:00
Jonas Mueller 78e8496363 Small tutorial notebook fixes/improvements (#420) 2022-09-09 17:36:27 +01:00
Hui Wen d8147caded Polish multiannotator docs (#422)
* docstring formatting

* add tiebreak info

* add example link to tutorial

* add missing backticks

* minor docstring text edits

* clarify "task"

Co-authored-by: Jonas Mueller <1390638+jwmueller@users.noreply.github.com>
2022-09-09 09:15:54 -07:00
Hui Wen 7416d9e714 update examples link (#421) 2022-09-09 09:09:21 -07:00
Jonas Mueller 006af01bad update migration guide to reflect versions >= 2.0 (#416)
* update migration guide to reflect >= 2.0

* sidebar updated
2022-09-08 22:09:23 -07:00
Jonas Mueller 62402d9ade move migration link into dropdown 2022-09-08 21:53:58 -07:00
Elías Snorrason 44cd416a96 add tutorial notebook to docs.cleanlab.ai (#411) 2022-09-08 18:27:28 -07:00
Elías Snorrason ce90af35c6 Review tutorial for token classification (#410)
* review tutorial for token classification

- update labels variable name in markdown block
- add newline for rendering bullet points
- put pulldown content into code blocks
- hide code blocks and move filepaths variable

* resolve pr comments

* remove pulldown code snippet and todo

* fix notebook json - delete extra comma

Co-authored-by: Jonas Mueller <1390638+jwmueller@users.noreply.github.com>
2022-09-08 18:27:11 -07:00
Hui Wen 1426a3c0fb add termcolor dependency (#415) 2022-09-08 12:59:01 -07:00
Jonas Mueller 23067aa9c6 Headers for newly added modules (#409)
* header for multiannotator module

* header for outlier module

* header for token_classification.filter

* header for token_classification.rank

* header for token_classification.summary

* header for util

* header for latent_algebra

* header for label_quality_utils

* header for token_classification_utils

* header for huggingfacekerasclassifier

* add experimental modules' dependencies to readme
2022-09-08 09:20:26 -07:00