Search papers, labs, and topics across Lattice.
The paper introduces the Lucie Training Dataset, a multilingual corpus emphasizing French and minimizing copyrighted material, and the Lucie-7B foundation model, trained on a balanced French-English dataset. They address the anglo-centric bias in LLM pretraining data by incorporating French cultural heritage documents and prioritizing data rights. The release of Lucie-7B and its instruction-tuned variants, along with training code and data processing details, demonstrates strong performance while adhering to open-source principles.
Forget LLaMA, Lucie-7B offers a strong, openly licensed alternative trained on a multilingual dataset prioritizing French and data rights, proving you don't need massive scale or questionable data to achieve impressive results.
We present both the Lucie Training Dataset and the Lucie-7B foundation model. The Lucie Training Dataset is a multilingual collection of textual corpora centered around French and designed to offset anglo-centric biases found in many datasets for large language model pretraining. Its French data is pulled not only from traditional web sources, but also from French cultural heritage documents, filling an important gap in modern datasets. Beyond French, which makes up the largest share of the data, we added documents to support several other European languages, including English, Spanish, German, and Italian. Apart from its value as a resource for French language and culture, an important feature of this dataset is that it prioritizes data rights by minimizing copyrighted material. In addition, building on the philosophy of past open projects, it is redistributed in the form used for training and its processing is described on Hugging Face and GitHub. The Lucie-7B foundation model is trained on equal amounts of data in French and English -- roughly 33% each -- in an effort to better represent cultural aspects of French-speaking communities. We also describe two instruction fine-tuned models, Lucie-7B-Instruct-v1.1 and Lucie-7B-Instruct-human-data, which we release as demonstrations of Lucie-7B in use. These models achieve promising results compared to state-of-the-art models, demonstrating that an open approach prioritizing data rights can still deliver strong performance. We see these models as an initial step toward developing more performant, aligned models in the near future. Model weights for Lucie-7B and the Lucie instruct models, along with intermediate checkpoints for the former, are published on Hugging Face, while model training and data preparation code is available on GitHub. This makes Lucie-7B one of the first OSI compliant language models according to the new OSI definition.
