Releases: deepchecks/deepchecks
Releases · deepchecks/deepchecks
0.18.1
0.18.0
Documentation
- fix typos in README.md [skip ci] (#2699) #2699 (Eyal C)
- add RayanAAY-ops as a contributor for code, and ideas (#2733) #2733 (allcontributors[bot])
Commits
- d1c07ab: Change nlp model properties to use onnx optimized models (Nadav Barak) #2681
- 3eaca24: Remove empty_gpu for cached models (Nadav Barak) #2682
- 9890fd5: Part of the vulnerability fixes required by Snyk (#2683) (Noam Bressler) #2683
- 3f148f1: Fix CI/CD (#2685) (Harsh Jain) #2685
- c2bbd59: Bump dev version (#2686) (Noam Bressler) #2686
- cf4a7ea: Added support for filterting check results by name (#2695) (Harsh Jain) #2695
- a40e8c3: modification of language property text (#2704) (Nadav Barak) #2704
- 9806488: Change lexical density to a 0-1 float (#2708) (Noam Bressler) #2708
- f8eaa0c: Vulnerability issues fix by Synk (#2703) (Harsh Jain) #2703
- 6ff7d2a: Code optimization for cleaning special chars from string (#2698) (Manish Kumar) #2698
- 76b92c8: change_weak_segments_na_logic (#2709) (JKL98ISR) #2709
- c4af7dd: segments bug fix with na (#2712) (Nadav Barak) #2712
- 708bac7: weak segment should have maximum of 1 category (#2705) (Nadav Barak) #2705
- 7bb31d1: remove ipython req (#2716) (Noam Bressler) #2716
- 42cf4b3: improved documentation (#2717) (Nadav Barak) #2717
- 7639a35: Limit a newer sklearn version that breaks _ProbaScorer (#2723) (Noam Bressler) #2723
- 2ef42a0: Fix build (#2731) (Noam Bressler) #2731
- 5db924d: Remove benchmark history check (#2732) (Noam Bressler) #2732
- 32cb218: update ver and docs (Noam Bressler)
0.17.5
Documentation
- add arterm-sedov as a contributor for doc (#2662) #2662 (allcontributors[bot])
Commits
- 5ac2107: bump version (Noam Bressler)
- 1c539f6: more robust nan handling when creating TextData label (#2653) (Noam Bressler) #2653
- e617151: switch to better Toxicty model (#2656) (Noam Bressler) #2656
- 2d0cbf5: Corrected grammar in dataset_object.rst (#2661) (arterm-sedov) #2661
- f9251b2: A new check - frequent substrings: (shiritdvir) #2654
- fa81d36: Adding a doc to the new check - frequent substring (#2663) (shiritdvir) #2663
- 11bc134: calibrate the toxicity model (#2664) (Noam Bressler) #2664
- f378c27: Slight adjustment to prefer larger segments (#2665) (Noam Bressler) #2665
- fd0e529: use device_map=auto when accelerate is installed (#2666) (Noam Bressler) #2666
- 853084d: remove extended (#2655) (Noam Bressler) #2655
- 7fd01a2: avoid installing future versions of pandas, and solve issue with pandas 2.1.0 (#2669) (Noam Bressler) #2669
- 1586dd0: cache models on cpu + hugging face based quantization (#2671) (Nadav Barak) #2671
- 0115c80: Update code owners (#2672) (Noam Bressler) #2672
- ee7e31a: Merge branch 'main' into 0.17.x (Noam Bressler)
0.17.4
Commits
- fefb941: Bump hotfix version (#2601) (Noam Bressler) #2601
- 300218b: [DOCS] Update README.md (#2602) (shir22) #2602
- dacf11c: Update drift guide for nlp (#2595) (Nir Hutnik) #2595
- f15f3be: multivariate drift fix (#2603) (JKL98ISR) #2603
- b1fa92e: Changed model_classes logger warning and made sure tabular only warns once (#2604) (Nir Hutnik) #2604
- c521fc7: fix many warnings when using joblib (weak segments) (#2609) (JKL98ISR) #2609
- 18aa661: Not raise DeepchecksProcessError for UnderAnnotated + WeakSegments (#2611) (Harsh Jain) #2611
- dac4bc6: Added .describe() function to the TextData class (#2606) (Harsh Jain) #2606
- 17455f5: Added is_english property & flag to calculate english-only properties on non-english identified samples (#2616) (Nir Hutnik) #2616
- 3ec7ac8: Update scipy version and broken pillow dependencies in requirements (#2620) (Harsh Jain) #2620
- e481a87: Improved heatmap display for confusion matrix report (#2617) (Harsh Jain) #2617
- 908e4da: Calculation of NLP properties with batches (#2621) (Itay Gabbay) #2621
- f3c420f: small refactor in properties calculation (#2627) (Nadav Barak) #2627
- d21213d: bug fix properties calc on single sample batch (#2628) (Nadav Barak) #2628
- 78eebd2: Add long description for properties, and update some of the short ones. (#2629) (Noam Bressler) #2629
- 6017cb9: fix nlp device kwargs (#2632) (JKL98ISR) #2632
- 2ea001e: Add examples to new and existing properties (#2634) (Noam Bressler) #2634
- 5723808: Amend docs - use spearman rather than Pearson (#2636) (Noam Bressler) #2636
- 3071ebf: Differentiate special chars from punctuations (#2635) (Noam Bressler) #2635
- 407652a: Rename Readability Score to Reading Ease (#2639) (Noam Bressler) #2639
- fa069e6: Fix (Itay Gabbay)
- bd95147: Add ability to cache text prop models (#2641) (Shay Tsadok) #2641
- de04894: Fix mixup in dataset names and colors in feature drift (#2644) (Noam Bressler) #2644
- deb16bc: Fix the Simple Model Comp condition (#2647) (Noam Bressler) #2647
- daad127: Weak segment - return single segment per feature argument (#2645) (Nadav Barak) #2645
- 79ad491: rename is english to english text (#2648) (Noam Bressler) #2648
- fcd084e: Return nan embeddings for empty samples (#2649) (Noam Bressler) #2649
- 3adfe0a: Merge remote-tracking branch 'origin/0.17.x' into 0.17.x (Noam Bressler)
- 5b0f2e2: Merge branch 'main' into 0.17.x (Noam Bressler)
0.17.3
Commits
- 79422c2: bump hotfix version (#2592) (Noam Bressler) #2592
- 7847336: [DOCS] Update Readme (#2593) (shir22) #2593
- e22250e: Fix new albumentations issue (#2594) (Nir Hutnik) #2594
- 867c934: [DOCS] Update Continuous Validation in Welcome (#2596) (shir22) #2596
- 057d795: Setting any non-numeric value to nan and tests (#2597) (Harsh Jain) #2597
- 62a73ae: airbnb fixes and improvements (#2600) (JKL98ISR) #2600
- f5c10d3: Merge branch 'main' into 0.17.x (Noam Bressler)
0.17.2
0.17.1
0.17.0
Documentation
- add hjain5164 as a contributor for doc (#2555) #2555 (allcontributors[bot])
- add hjain5164 as a contributor for bug (#2564) #2564 (allcontributors[bot])
Commits
- f728941: update version to 0.14.1.dev1 (#2546) (Noam Bressler) #2546
- d84ac71: [DEE-550] fix - 'dataset.select' mutates provided columns list (#2544) (Yurii Romanyshyn) #2544
- c153794: updated embeddings calculation functions names (#2548) (Nir Hutnik) #2548
- 5654063: fix deprecated styler method (#2550) (matanper) #2550
- 87de7d7: Noam/bugfix/token class bugfixes (#2551) (Noam Bressler) #2551
- a9c91cf: Fix prediction/label drift for token classification (#2553) (Nir Hutnik) #2553
- b097a88: Elaborate contribution guide for windows (Harsh Jain) #2534
- a4144b3: Fixed message for missing embeddings (#2554) (Nir Hutnik) #2554
- 1ab9aef: Fixed minor bug in TrainTestPerformance condition (Nir Hutnik) #2556
- ce15432: iqr range modification (#2557) (Nadav Barak) #2557
- 7c031aa: Add min_samples and n_top_classes to TrainTestPerformance in NLP (#2558) (Nir Hutnik) #2558
- bc74006: Missing package name in NLP quickstart (#2563) (Harsh Jain) #2563
- a487b0c: Nb/bug/weak segments fill na (#2560) (Nadav Barak) #2560
- cd1898f: data duplicates takes a long time for long docs (#2561) (Noam Bressler) #2561
- 7d311f5: Create token classification use case (#2559) (Noam Bressler) #2559
- 269a170: fix error in pandas 2 (#2567) (matanper) #2567
- 746a8c5: reorder checks in suites NLP (#2568) (Nadav Barak) #2568
- 7c5f3fb: Utilize fast tokenizers in Unknown Tokens check (Noam Bressler) #2569
- 133d395: Fix bug in binary classification support in nlp (Nadav Barak) #2571
- 70d6289: Improved run times of properties (Nir Hutnik) #2566
- 96c67c0: Fix the property drift check name and others various fixes (#2572) (Noam Bressler) #2572
- 70886a5: [MON-2340] moved airbnb files to s3 (#2575) (Yurii Romanyshyn) #2575
- 90cc7d6: Nb/feat/special chars check (#2576) (Nadav Barak) #2576
- ab64c80: Added description and URL for text properties in plots (#2574) (Harsh Jain) #2574
- 6889a3c: [DOCS] Update readme (unify testing, ci & testing management, monitoring) (#2577) (shir22) #2577
- 42bd8c6: [DOCS] Update Readme (#2580) (shir22) #2580
- ac7b35d: Embed long documents with openai (#2573) (Noam Bressler) #2573
- 5a39ec5: take properties type from built in dict and not from user. (#2579) (Noam Bressler) #2579
- d1af76b: replace airbnb link (#2584) (matanper) #2584
- 59464b8: condition for outlier check + bug fix in validation (Nadav Barak) #2583
- b62a8bd: Noam/address null properties (#2585) (Noam Bressler) #2585
- 43dc418: Fix bugs in properties - make long run calcs work faster again. (Nir Hutnik) #2581
- 1e24c21: updated version (nirhutnik)
0.14.1
Commits
- 6b6c4cb: [DEE-456] nlp dummy model refactoring (#2511) (Yurii Romanyshyn) #2511
- 90d34eb: update main to 0.14.0 (#2521) (Noam Bressler) #2521
- fe3f69e: Plot files under annotated checks (#2516) (Nadav Barak) #2516
- 3ebf01c: Fixes #2454 Added Readability score and average sentence length text property (#2512) (Harsh Jain) #2512
- 63606de: fix version in docs (#2522) (JKL98ISR) #2522
- 3ff8209: fix version in docs (#2522) (JKL98ISR)
- f205e23: change nlp to beta (#2527) (Noam Bressler) #2527
- d6bab6a: Changed text embeddings to numpy array (Nir Hutnik) #2528
- 78124a9: Nb/feat/support multi label (#2531) (Nadav Barak) #2531
- a08cb72: Added support of token classification to PredictionDrift & LabelDrift (#2533) (Nir Hutnik) #2533
- 7f88992: Updated label display in NLP (#2532) (Nir Hutnik) #2532
- 90fff26: [DEE-417] do not calculate english-only text properties for not english samples (#2525) (Yurii Romanyshyn) #2525
- a2b9fce: [DOCS] Add JMLR Paper reference to CITATION.cff (#2486) (shir22) #2486
- 32fb463: Extended just dance dataset + small fixes (#2538) (Nadav Barak) #2538
- cdab603: Remove langdetect from requirements (Nir Hutnik) #2535
- aaafe51: Noam/dee 593 plot example for integrity checks (#2542) (Noam Bressler) #2542
- 1baf39f: Noam/dee 530 feat improved quickstart for nlp (#2540) (Noam Bressler) #2540
- cac68b9: fix set_precision in pandas deprecated (#2543) (matanper) #2543
- a886103: multi label quick start (#2541) (Nadav Barak) #2541
- a1f921e: Fixes #2524 Refactoring of the calculate_default_properties function and adding new text properties (#2536) (Harsh Jain) #2536
- fbecc3e: fix reference to nlp in vision (#2545) (Noam Bressler) #2545
- 119104d: Merge branch 'main' into 0.14.x (Noam Bressler)
0.14.0
Documentation
- add thewchan as a contributor for code (#2441) #2441 (allcontributors[bot])
- add hjain5164 as a contributor for code (#2499) #2499 (allcontributors[bot])
Commits
- 2a24338: updated dataset - airbnb (#2434) (Nadav Barak) #2434
- 656d5fc: refactor text data (#2432) (Nadav Barak) #2432
- 1657295: Add Additional Permissions to License (shir22) #2436
- 9449798: Improve column type inference (#2440) (matanper) #2440
- d246a34: fix sklearn warning (#2442) (matanper) #2442
- 32536c8: Update MANIFEST.in to include nlp-prop-requirements.txt (#2437) (Matt Chan) #2437
- a084852: [DEE-443;DEE-460] text properties refactoring; removal of 'TextData.properties_types' and 'TextData.metadata_types' (#2438) (Yurii Romanyshyn) #2438
- 4dc1bad: nlp doc structure + metrics_guide.rst (#2449) (Nadav Barak) #2449
- 78867f8: don't fail on langdetect (#2444) (matanper) #2444
- c03a2ae: tweet eval ds update + single dataset performance (#2451) (Nadav Barak) #2451
- 0d105c8: [DEE-440] added error msg to checks that do not support token/multi-label classification (#2445) (Yurii Romanyshyn) #2445
- 76d5957: performance docs NLP (#2453) (Nadav Barak) #2453
- 750d2d6: [DEE-457] "PropertyDrift" and "TrainTestPerformance" checks (#2456) (Yurii Romanyshyn) #2456
- ca57913: update_version_to_comply_with_new_pep (#2466) (JKL98ISR) #2466
- 53c8d87: fix hide index (#2465) (matanper) #2465
- 1bec6b5: Nb/feat/multi label (#2467) (Nadav Barak) #2467
- aee73d1: [DEE-486] nlp 'token_*_per_class' scorers fix (#2458) (Yurii Romanyshyn) #2458
- e4429e3: documentation restructure (Nadav Barak) #2457
- 0b39ef1: Added properties and metadata guides (#2468) (Nir Hutnik) #2468
- cfb005c: Remove failing test (#2473) (Noam Bressler) #2473
- 019ac48: [DEE-510] nlp integrity checks (#2471) (Yurii Romanyshyn) #2471
- c733c14: Examples for property drift and train test performance (Noam Bressler) #2470
- c565c04: update_cli_ubuntu (#2474) (JKL98ISR) #2474
- c503c94: Add unknown token check (#2483) (Noam Bressler) #2483
- c369227: [DOCS] Update Main Page for New Structure (#2484) (shir22) #2484
- d99add3: Nb/docs/warnings fix (#2482) (Nadav Barak) #2482
- 3bbaa09: simplify weak segments (#2485) (Nadav Barak) #2485
- 2e08fd8: Update welcome and add link to Deepchecks Monitoring (#2490) (shir22) #2490
- bcbfa68: add test data docs (#2488) (Noam Bressler) #2488
- 294ed16: Noam/dee 538 better handle download of nltk (#2493) (Noam Bressler) #2493
- 728cc44: Weak segments refactor + bug fix (#2491) (Nadav Barak) #2491
- 7f107b2: [BUG] added forgotten 'raise' keyword (#2496) (Yurii Romanyshyn) #2496
- 4a66c72: Fixes #2452 Added two new text properties (#2478) (Harsh Jain) #2478
- 0e011b0: Several fixes (Noam Bressler) #2494
- f9bf86d: Noam/bugfix/nlp various fixes (Noam Bressler) #2501
- 0138592: [DEE-506] added 'models_storage' param to the 'calculate_default_properties' function (#2477) (Yurii Romanyshyn) #2477
- 9d1261c: Feat/Add Misclassified samples lower than condition to Confusion Matrix Report (#2500) (Kishore Sampath) #2500
- 122263b: Improve lang detection (#2506) (Noam Bressler) #2506
- 5e60e95: Noam/dee 480 support weak segments in multilabel (#2507) (Noam Bressler) #2507
- 7e0365b: Fix api docs (#2508) (Itay Gabbay) #2508
- 8e73fe2: Noam/feature/update default suites (#2510) (Noam Bressler) #2510
- 45bc883: Nb/feat/under annotated checks (#2505) (Nadav Barak) #2505
- 56bedfb: Added embeddings drift check + embeddings support in TextData (#2502) (Nir Hutnik) #2502
- 411a25a: under annotated check - support multi label and token classification (#2513) (Nadav Barak) #2513
- 0dc9d7f: [DOCS] Update ci-cd order and fix link (#2514) (shir22) #2514
- 59535f3: small fixes (#2515) (Noam Bressler) #2515
- 7b49dc2: 0.14.0 release (Noam Bressler)
- cfc9f9d: fix fasttext file path issue (#2518) (Noam Bressler) #2518
- 03960dc: Merge branch 'main' into 0.14.x (Noam Bressler)