distilabel
distilabel copied to clipboard
argilla-io
Distilabel is a framework for synthetic data and AI feedback for engineers who need fast, reliable and scalable pipelines based on verified research papers.
## Description This PR adds a base method to allow easily printing the prompts of the tasks. ```python from distilabel.steps.tasks import URIAL # only importable if placed at the root...
**Is your feature request related to a problem? Please describe.** It would be nice to intergrate https://llama-cpp-python.readthedocs.io/en/stable/#embeddings because of the speed of default `sentence-transformers` **Describe the solution you'd like** similar...
## Description This PR adds [CLAIR](https://arxiv.org/abs/2408.06266v1) task to revise instructions: *CLAIR uses an AI system to minimally revise a solution A→A´ such that the resulting preference A `preferred` A’ is...
MagpieGenerator now is not supports LlamaCppLLM. I think it will be useful feature.
**Describe the bug** `TextClassification` only generates structured text, which can result on errors hard to get. **To Reproduce** Code to reproduce ```python ``` **Expected behaviour** Work as any other task...
**Describe the bug** There's an integration test failing in the CI due to a TimeOut apparently, sometimes a python version, sometimes another one. An example can be seen in this...
## Which page or section is this issue related to? The new steps for text clustering would benefit from a example entry in the docs: `UMAP`, `DBSCAN` and `TextClustering` (and...
**Is your feature request related to a problem? Please describe.** The components gallery show the steps all at once, but having them grouped by the icons we already generate could...
**Is your feature request related to a problem? Please describe.** - where can I find the system prompt? - how easy is it to overwrite/adapt the prompt template? - can...
## Which page or section is this issue related to? people seem to be unaware of the structured generation feature ## What are you documenting, or what change are you...
Metadata
Owner
Metadata
Distilabel is a framework for synthetic data and AI feedback for engineers who need fast, reliable and scalable pipelines based on verified research papers.