1.16.1

Bug fixes

• Fix import datasets on python 3.10 by @lhoestq in #3326
• Fix wrongly converted assert by @eliasws in #3323

1.16.0

Datasets Changes

• New: riddle_sense by @ziyiwu9494 in #3161
• New: Multi-Lingual LibriSpeech by @patrickvonplaten in #3198
• New: XCSR by @yangxqiao in #3074
• New: CMU Hinglish DoG by @Ishan-Kumar2 in #3149
• New: Multidoc2dial by @sivasankalpp in #3205
• New: IndoNLI by @afaji in #3307
• Update: DaNE - updated URL for download by @MalteHB in #3203
• Update: xcopa - (fix checksum issues + add translated data) by @mariosasko in #3254
• Update: tatoeba - update to v2021-07-22 by @KoichiYasuoka in #3225
• Update: KILT - update metadata JSON by @albertvillanova in #3276
• Update: Covost 2 - update download instructions by @patrickvonplaten in #3281
• Update: Common Voice, OpenSLR, LibriSpeech ASR, Vivos - make several audio datasets streamable by @lhoestq in #3290
• Fix: tuple_ie - fix download url by @mariosasko in #3213
• Fix: id_newspapers_2018 - fix streaming by @lhoestq in #3249
• Fix: bookcorpusopen - fix RAM usage by @lhoestq in #3280
• Fix: Scielo - fix ConnectionError by @mariosasko in #3260
• Fix: tatoeba - fix URLs for a subset of xtreme by @mariosasko in #3321

Datasets Features

• Push to hub capabilities for Dataset and DatasetDict by @LysandreJik in #3098:
  • upload your dataset to the Hugging face Hub with the push_to_hub() method !
  • See documentation here
• 200+ datasets now support streaming:
  • Stream TAR-based dataset using iter_archive by @lhoestq in #3110
  • Stream from Google Drive and other hosts by @lhoestq in #3248
  • Support Audio feature in streaming mode by @albertvillanova in #3133
  • Support Audio feature for TAR archives in sequential access by @albertvillanova in #3129
• Resolve data_files by split name automatically by @lhoestq in #3221
  • It takes into account the file names to know which file goes into which split
  • See documentation here
• Filter method for batched=True by @thomasw21 in #3244
• Adding with_rank arg to pass process rank to map by @TevenLeScao in #3314

Dataset Cards

• Add full tagset to conll2003 README by @BramVanroy in #3230
• Fix some contact information formats by @lhoestq in #3274
• Add wikipedia tags by @lhoestq in #3301
• Updating details of IRC disentanglement data by @jkkummerfeld in #3259

Metrics Changes

• New: OpenAI's pass@k code evaluation metric by @lvwerra in #2916
• Update: BLEURT - options to use updated bleurt checkpoints by @jaehlee in #3235
• Update: CER - update to support latest release by @mariosasko in #3252
• Update: WER - update to the documentation by @wooters in #3278

Documentation

• Add docs for to_tf_dataset by @stevhliu in #3175
• Small updates to to_tf_dataset documentation by @Rocketknight1 in #3215
• Update link to Datasets Tagging app in Spaces by @albertvillanova in #3194
• Improve repository structure docs by @lhoestq in #3233
• Swap descriptions of v1 and raw-v1 configs of WikiText dataset and fix metadata by @albertvillanova in #3241
• Add docs for audio processing by @stevhliu in #3222
• Add push_to_hub docs by @lhoestq in #3319

Additional improvements and bug fixes

• Catch token invalid error in CI by @lhoestq in #3200
• Pin keras version until TF fixes its release by @albertvillanova in #3208
• Fix disable_nullable default value to False by @lhoestq in #3211
• Fix code quality in riddle_sense dataset by @albertvillanova in #3218
• Better error msg if len(predictions) doesn't match len(references) in metrics by @mariosasko in #3160
• Use huggingface_hub.HfApi to list datasets/metrics by @mariosasko in #3121
• Pin version exclusion for tensorflow incompatible with keras by @albertvillanova in #3216
• Group tests in multiprocessing workers by test file by @albertvillanova in #3231
• Fix load_from_disk temporary directory by @lhoestq in #3245
• [tiny] fix typo in stream docs by @nollied in #3246
• Avoid PyArrow type optimization if it fails by @mariosasko in #3234
• Remove redundant isort module placement by @mariosasko in #3243
• asserts replaced by exception for text classification task with test. by @manisnesan in #3256
• Add os.listdir for streaming by @lhoestq in #3270
• asserts replaced with exception for image classification task, csv, json by @manisnesan in #3262
• Force data files extraction if download_mode='force_redownload' by @mariosasko in #3275
• Minor Typo Fix - Precision to Recall by @SebastinSanty in #3279
• Decode audio from remote by @lhoestq in #3271
• Fix build_docs CI by @lhoestq in #3286
• Allow datasets with indices table when concatenating along axis=1 by @mariosasko in #3288
• f-string formatting by @Mehdi2402 in #3277
• Unpin markdown for build_docs now that it's fixed by @lhoestq in #3289
• Pin version exclusion for Markdown by @albertvillanova in #3293
• Use f-strings in the dataset scripts by @Carlosbogo in #3291
• fix old_val typo in f-string by @Mehdi2402 in #3302
• asserts replaced with exception for fingerprint.py, search.py, arrow_writer.py and metric.py by @Ishan-Kumar2 in #3305
• fix: files counted twice in inferred structure by @borisdayma in #3309
• Finish transition to PyArrow 3.0.0 by @mariosasko in #3318
• Removing query params for dynamic URL caching by @anton-l in #3315

Citation

• Update BibTeX entry by @albertvillanova in #3223
• Fix paper BibTeX citation with proceedings reference by @albertvillanova in #3226
• Add CITATION file by @albertvillanova in #3228
• Fix URL in CITATION file by @albertvillanova in #3229

Deprecations

• Deprecate prepare_module by @albertvillanova in #3166

Full Changelog: 1.15.1...1.16.0

1.15.1

Dependencies

• Bump huggingface_hub to 0.1.0 by @lhoestq in #3199

1.15.0

Dataset Changes

• Update: JNLBA - add tags names by @bhavitvyamalik in #3092
• Update: OpenSLR - add SLR83 to OpenSLR by @tyrius02 in #3125 and #3176
• Update: RONEC - update to v2 by @dumitrescustefan in #3184
• Fix: Arabic Billion Words - Fix script to return all data by @albertvillanova in #3136
• Fix: HLGD - fix label mapping by @VictorSanh in #3180

Dataset Features

• Allow dynamic first dimension for ArrayXD by @rpowalski in #2891
• add multi-proc in to_csv by @bhavitvyamalik in #2896
• QOL improvements: auto-flatten_indices and desc in map calls by @mariosasko in #3196

Dataset Cards

• Fill in dataset card for NCBI disease dataset by @edugp in #3115

Metrics Changes

• New: metric for the MATH dataset (competition_math). by @hacobe in #3020
• New: Google BLEU (aka GLEU) metric by @slowwavesleep in #3108
• New: TER by @BramVanroy in #3153
• New: ChrF(++) by @BramVanroy in #3187

General improvements and bug fixes

• Correctly update metadata to preserve features when concatenating datasets with axis=1 by @mariosasko in #3120
• Fixes to to_tf_dataset by @Rocketknight1 in #3085
• Add security policy to the project by @albertvillanova in #2958
• Update doc links to point to new docs by @mariosasko in #3116
• Fix caching bugs by @mariosasko in #3141
• Fix numpy deprecation warning for ragged tensors by @lhoestq in #3137
• Fixed: duplicate parameter and missing parameter in docstring by @PanQiWei in #3157
• Fix some typos in the documentation by @h4iku in #3152
• Fix string encoding for Value type by @lhoestq in #3158
• Fix CLI test to ignore verfications when saving infos by @albertvillanova in #3147
• Make inspect.get_dataset_config_names always return a non-empty list by @albertvillanova in #3159
• Fix issue with filelock filename being too long on encrypted filesystems by @mariosasko in #3173
• Asserts replaced by exceptions (#3171) by @joseporiolayats in #3174
• Preserve ordering in zip_dict by @mariosasko in #3170
• Don't memoize strings when hashing since two identical strings may have different python ids by @lhoestq in #3182
• Re-add faiss to windows testing suite by @BramVanroy in #3151
• Add missing docstring to DownloadConfig by @mariosasko in #3183
• More efficient nested features encoding by @eladsegal in #3124
• Fix optimized encoding for arrays by @lhoestq in #3197

1.14.0

Dataset changes

• Update: LexGLUE and MultiEURLEX README - update dataset cards #3075 (@iliaschalkidis)
• Update: SUPERB - use Audio features #3101 (@anton-l)
• Fix: Blog Authorship Corpus - fix URLs #3106 (@albertvillanova)

Dataset features

• Add iter_archive #3066 (@lhoestq)

General improvements and bug fixes

• Replace FSTimeoutError with parent TimeoutError #3100 (@albertvillanova)
• Fix project description in PyPI #3103 (@albertvillanova)
• Align tqdm control with cache control #3031 (@mariosasko)
• Add paper BibTeX citation #3107 (@albertvillanova)

1.13.3

Dataset changes

• Update: Adapt all audio datasets #3081 (@patrickvonplaten)

Bug fixes

• Update BibTeX entry #3090 (@albertvillanova)
• Use template column_mapping to transmit_format instead of template features #3088 (@mariosasko)
• Fix Audio feature mp3 resampling #3096 (@albertvillanova)

1.13.2

Bug fixes

• Fix error related to huggingface_hub timeout parameter #3082 (@albertvillanova)
• Remove _resampler from Audio fields #3086 (@albertvillanova)

1.13.1

Bug fixes

• Fix loading a metric with internal import #3077 (@albertvillanova)

1.13.0

Dataset changes

• New: CaSiNo #2867 (@kushalchawla)
• New: Mostly Basic Python Problems #2893 (@lvwerra)
• New: OpenAI's HumanEval #2897 (@lvwerra)
• New: SemEval-2018 Task 1: Affect in Tweets #2745 (@maxpel)
• New: SEDE #2942 (@Hazoom)
• New: Jigsaw unintended Bias #2935 (@Iwontbecreative)
• New: AMI #2853 (@cahya-wirawan)
• New: Math Aptitude Test of Heuristics #2982 #3014 (@hacobe, @albertvillanova)
• New: SwissJudgmentPrediction #2983 (@JoelNiklaus)
• New: KanHope #2985 (@adeepH)
• New: CommonLanguage #2989 #3006 #3003 (@anton-l, @albertvillanova, @jimregan)
• New: SwedMedNER #2940 (@bwang482)
• New: SberQuAD #3039 (@Alenush)
• New: LexGLUE: A Benchmark Dataset for Legal Language Understanding in English #3004 (@iliaschalkidis)
• New: Greek Legal Code #2966 (@christospi)
• New: Story Cloze Test #3067 (@zaidalyafeai)
• Update: SUPERB - add IC, SI, ER tasks #2884 #3009 (@anton-l, @albertvillanova)
• Update: MENYO-20k - repo has moved, updating URL #2939 (@cdleong)
• Update: TriviaQA - add web and wiki config #2949 (@shirte)
• Update: nq_open - Use standard open-domain validation split #3029 (@craffel)
• Update: MeDAL - Add further description and update download URL #3022 (@xhlulu)
• Update: Biosses - fix column names #3054 (@bwang482)
• Fix: scitldr - fix minor URL format #2948 (@albertvillanova)
• Fix: masakhaner - update JSON metadata #2973 (@albertvillanova)
• Fix: TriviaQA - fix unfiltered subset #2995 (@lhoestq)
• Fix: TriviaQA - set writer batch size #2999 (@lhoestq)
• Fix: LJ Speech - fix Windows paths #3016 (@albertvillanova)
• Fix: MedDialog - update metadata JSON #3046 (@albertvillanova)

Metric changes

• Update: meteor - update from nltk update #2946 (@lhoestq)
• Update: accuracy,f1,glue,indic-glue,pearsonr,prcision,recall-super_glue - Replace item with float in metrics #3012 #3001 (@albertvillanova, @mariosasko)
• Fix: f1/precision/recall metrics with None average #3008 #2992 (@albertvillanova)
• Fix meteor metric for version >= 3.6.4 #3056 (@albertvillanova)

Dataset features

• Use with TensorFlow:
  • Adding to_tf_dataset method #2731 #2931 #2951 #2974 (@Rocketknight1)
• Better support for ZIP files:
  • Support loading dataset from multiple zipped CSV data files #3021 (@albertvillanova)
  • Load private data files + use glob on ZIP archives for json/csv/etc. module inference #3041 (@lhoestq)
• Streaming improvements:
  • Extend support for streaming datasets that use glob.glob #3015 (@albertvillanova)
  • Add remove_columns to IterableDataset #3030 (@cccntu)
  • All the above ZIP features also work in streaming mode
• New utilities:
  • Add get_dataset_split_names() to get a dataset config's split names #2906 (@severo)
• Replace script_version with revision #2933 (@albertvillanova)
  • The script_version parameter in load_dataset is now deprecated, in favor of revision
• Experimental - Create Audio feature type #2324 (@albertvillanova):
  • It allows to automatically decode audio data (mp3, wav, flac, etc.) when examples are accessed

Dataset cards

• Add arxiv paper inswiss_judgment_prediction dataset card #3026 (@JoelNiklaus)

Documentation

• Add tutorial for no-code dataset upload #2925 (@stevhliu)

General improvements and bug fixes

• Fix filter leaking #3019 (@lhoestq)
  • calling filter several times in a row was not returning the right results in 1.12.0 and 1.12.1
• Update BibTeX entry #2928 (@albertvillanova)
• Fix exception chaining #2911 (@albertvillanova)
• Add regression test for null Sequence #2929 (@albertvillanova)
• Don't use old, incompatible cache for the new filter #2947 (@lhoestq)
• Fix fn kwargs in filter #2950 (@lhoestq)
• Use pyarrow.Table.replace_schema_metadata instead of pyarrow.Table.cast #2895 (@arsarabi)
• Check that array is not Float as nan != nan #2936 (@Iwontbecreative)
• Fix missing conda deps #2952 (@lhoestq)
• Update legacy Python image for CI tests in Linux #2955 (@albertvillanova)
• Support pandas 1.3 new read_csv parameters #2960 (@SBrandeis)
• Fix CI doc build #2961 (@albertvillanova)
• Run tests in parallel #2954 (@albertvillanova)
• Ignore dummy folder and dataset_infos.json #2975 (@Ishan-Kumar2)
• Take namespace into account in caching #2938 (@lhoestq)
• Make Dataset.map accept list of np.array #2990 (@albertvillanova)
• Fix loading compressed CSV without streaming #2994 (@albertvillanova)
• Fix json loader when conversion not implemented #3000 (@lhoestq)
• Remove all query parameters when extracting protocol #2996 (@albertvillanova)
• Correct a typo #3007 (@Yann21)
• Fix Windows test suite #3025 (@albertvillanova)
• Remove unused parameter in xdirname #3017 (@albertvillanova)
• Properly install ruamel-yaml for windows CI #3028 (@lhoestq)
• Fix typo #3023 (@qqaatw)
• Extend support for streaming datasets that use glob.glob #3015 (@albertvillanova)
• Actual "proper" install of ruamel.yaml in the windows CI #3033 (@lhoestq)
• Use cache folder for lockfile #2887 (@Dref360)
• Fix streaming: catch Timeout error #3050 (@borisdayma)
• Refac module factory + avoid etag requests for hub datasets #2986 (@lhoestq)
• Fix task reloading from cache #3059 (@lhoestq)
• Fix test command after refac #3065 (@lhoestq)
• Fix Windows CI with FileNotFoundError when setting up s3_base fixture #3070 (@albertvillanova)
• Update summary on PyPi beyond NLP #3062 (@thomwolf)
• Remove a reference to the open Arrow file when deleting a TF dataset created with to_tf_dataset #3002 (@mariosasko)
• feat: increase streaming retry config #3068 (@borisdayma)
• Fix pathlib patches for streaming #3072 (@lhoestq)

Breaking changes:

• Due to the big refactoring at #2986, the prepare_module function doesn't support the return_resolved_file_path and return_associated_base_path parameters. As an alternative, you may use the dataset_module_factory instead.

1.12.1

Bug fixes

• Fix fsspec AbstractFileSystem access #2915 (@pierre-godard)
• Fix unwanted tqdm bar when accessing examples #2920 (@lhoestq)
• Fix conversion of multidim arrays in list to arrow #2922 (@lhoestq):
  • this fixes the ArrowInvalid: Can only convert 1-dimensional array values errors