1.5.0
What's Changed
- Add common typing module by @plaguss in #1029
- docs: textcat tutorial by @sdiazlor in #949
- Add
task
decorator by @gabrielmbmb in #1028 - Update
docs
workflows to useuv
by @gabrielmbmb in #1032 - fix: simplify prompt template
ArgillaLabeller
by @davidberenstein1957 in #1033 - Add
dataset_batch_size
argument by @gabrielmbmb in #1039 - Move all LLMs to distilabel.models by @plaguss in #1045
- Fix a tiny typo in
_Step
docstring by @sadra-barikbin in #1051 - docs: improve docs for
MinHashDedup
Step
by @anakin87 in #1050 - Fix new response_format variable in openai api by @plaguss in #1053
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #1043
- Update
LLM.generate
output to includestatistics
by @plaguss in #1034 - Add example of structured output. by @plaguss in #1061
- feat: implenent basic SFT pipeline based on synthetic data generator by @burtenshaw in #1059
- fix: broken import in instruction by @burtenshaw in #1063
- Fix StepOutput type by @plaguss in #1072
- docs: update issue templates by @sdiazlor in #1074
- Update
unload
method fromvLLM
to properly free resources by @gabrielmbmb in #1077 - Add tasks to replicate Math-shepherd by @plaguss in #1052
- Add
load_groups
argument torun
by @gabrielmbmb in #1075 - Add
TextGenerationWithImage
task by @plaguss in #1066 - Create columns with
LLM
returned extra keys by @gabrielmbmb in #1078 - Fix
vLLM
unload logic when model isNone
by @gabrielmbmb in #1080 - Fix
merge_distilabel_metadata
function when handling outputs fromTask
withgroup_generations==True
by @gabrielmbmb in #1082 - chore: update base.py by @eltociear in #1085
- Add magpie support llama cpp ollama by @davidberenstein1957 in #1086
- Feat/954 llama cpp by @bikash119 in #1000
- fix import by replacing GeneratorOutput with GeneratorStepOutput by @davidberenstein1957 in #1093
- add mlx support by @davidberenstein1957 in #1089
- Support custom default headers in
OpenAILLM
class. by @khulaifi95 in #1088 - fix/pip install messages by @davidberenstein1957 in #1095
- Fix handling empty list statistics by @gabrielmbmb in #1094
- update to outlines010 by @davidberenstein1957 in #1092
- update: search by match by @sdiazlor in #1096
- Add Legend to Component Gallery Icons by @ParagEkbote in #1090
- Image Language Models and
ImageGeneration
task by @plaguss in #1060 - Update
LLM
s to support prompt logprobs use-case by @gabrielmbmb in #1099 1.5.0
by @gabrielmbmb in #1100
New Contributors
- @sadra-barikbin made their first contribution in #1051
- @anakin87 made their first contribution in #1050
- @pre-commit-ci made their first contribution in #1043
- @eltociear made their first contribution in #1085
- @bikash119 made their first contribution in #1000
- @khulaifi95 made their first contribution in #1088
- @ParagEkbote made their first contribution in #1090
Full Changelog: 1.4.2...1.5.0