Dataset · argilla
Curated multilingual Dolly instructions
argilla/databricks-dolly-15k-curated-multilingual
An edited version of Databricks Dolly instructions in Spanish, German, and French.
What's inside
A text dataset ranging from 10 to 100 thousand examples in Spanish, German, and French. It is an improved version of the original Dolly v2 dataset, which underwent programmatic cleaning and manual moderation using Argilla tools (approximately 400 records corrected manually).
What you can do with it
- SFT fine-tuning of multilingual LLMs for instruction following in European languages
- Training LoRA adapters for language model localization
- Comparing generation quality on curated multilingual prompts
- Tasks
- Text generation
- Languages
- Spanish, German, French
- Size
- 10K–100K rows
- Downloads this month
- 668
- Likes
- 55
Models trained on it
- TheBloke/Llama2-70B-OASST-SFT-v10-GGUFText generation 2.7K
- OpenLLM-France/Lucie-7B-Instruct-human-dataText generation 1.4K
- mradermacher/Lucie-7B-Instruct-human-data-GGUF 357
- vllm-sr/Decision-1.0-Route-0.6BZero shot classification 103
- OpenAssistant/llama2-70b-oasst-sft-v10Text generation 87
- ITLL/Decision-1.0-Route-0.6BZero shot classification 38
- TheBloke/Llama2-70B-OASST-SFT-v10-GPTQText generation 34
- alexgusevski/Lucie-7B-Instruct-human-data-q6-mlxText generation 23
How to load it
from datasets import load_dataset
ds = load_dataset("argilla/databricks-dolly-15k-curated-multilingual")