1.3.0
What's Changed
- Add new step
CombineKeysby @plaguss in #747 - Refactor naming columns steps combinecolumns combinekeys expandcolumns by @davidberenstein1957 in #758
- Drop remove deprecated
LoadHubDatasetby @davidberenstein1957 in #759 - Add
requirementslist forPipelineby @plaguss in #720 - Add
StepResourcesand step replicas inPipelineby @gabrielmbmb in #750 - Add load stages by @gabrielmbmb in #760
- Update min required version to
python==3.9by @gabrielmbmb in #770 - Optionally include the pipeline script in the hub when pushing your distiset by @plaguss in #762
- Add
docs-pr.ymlanddocs-pr-close.ymlworkflows by @gabrielmbmb in #774 - Add
RayPipelineclass by @gabrielmbmb in #769 - Fixed closed PR workflow by @gabrielmbmb in #776
- Add
MagpieandMagpieGeneratortasks by @gabrielmbmb in #778 - Fix some issues related to
Magpietask by @gabrielmbmb in #783 - Add
end_with_userandinclude_system_promptflags toMagpietasks and handleNones. by @gabrielmbmb in #784 - Add workflow concurrency group for publishing docs by @gabrielmbmb in #796
- Add
_desired_num_gpusattribute toCudaDevicePlacementMixinby @gabrielmbmb in #795 - Compatibility with
vLLMwithtensor_parallel_sizeargument by @gabrielmbmb in #805 - Update default names in
GroupColumnsby @plaguss in #808 - Request batches to
GeneratorStepif only step in pipeline by @gabrielmbmb in #828 - Add default name for a pipeline by @plaguss in #809
- Update distilabel phrasing based on PR hugging face hub by @davidberenstein1957 in #821
- Some more
Magpieimprovements by @gabrielmbmb in #833 - Add
Embeddingsbase class,SentenceTransformerEmbeddingsclass,EmbeddingGenerationandFaissNearestNeighboursteps by @gabrielmbmb in #830 - Create file per hostname in
CudaDevicePlacementMixinby @gabrielmbmb in #814 - Create a
GeneratorStepfrom a dataset using a helper function by @plaguss in #812 - Do not take into account
disable_cuda_device_placementfor pipeline signature by @gabrielmbmb in #838 - Add
RewardModelScorestep by @gabrielmbmb in #840 - Fix
LoadDataFromHubattribute_datasethadellipsisby default instead ofNoneby @gabrielmbmb in #841 - Create
PlacementGroupfor steps usingvLLMby @gabrielmbmb in #842 - Update
argillaintegration to useargilla_sdkv2 by @alvarobartt in #705 - Make
overall-ratingthe default aspect forUltraFeedbacktask by @gabrielmbmb in #843 - fix typo index.md by @franperic in #844
- Use
CudaDevicePlacementMixininRewardModelScorestep by @gabrielmbmb in #845 - Gather GPUs per Ray node to create placement groups by @gabrielmbmb in #848
- Fix typo in docs by @plaguss in #850
- Add
xfailrouting batch function tests by @gabrielmbmb in #852 - Fix creating placement group when
pipeline_parallel_size>1by @gabrielmbmb in #851 - docs: 846 docs include google analytics by @davidberenstein1957 in #847
- Add
ClientvLLMclass by @gabrielmbmb in #854 - Add hard-negative flag to include similar challenging negatives on triplets by @plaguss in #856
- Add bibtex references in the docstrings to be shown in the README by @plaguss in #855
- distilabel
1.3.0by @gabrielmbmb in #857
New Contributors
- @franperic made their first contribution in #844
Full Changelog: 1.2.4...1.3.0