Loubna Allal
Machine Learning Engineer at Hugging Face
Paris, Ile-de-France
Summary
🤩
Rockstar🎓
Top SchoolLoubna Allal is a Machine Learning Engineer at Hugging Face with seven years of experience building and optimizing training pipelines for language models, including leading the development of SmolLM series. She combines deep academic foundations in mathematics and machine learning (ENS Paris-Saclay, Mines Nancy) with hands-on contributions to high-profile open-source projects like transformers and BigCode’s evaluation harness, improving checkpointing, generation, and evaluation tooling for code models. Her background spans research in PDEs and Hawkes processes to production-facing ML engineering, giving her a rare blend of theoretical rigor and practical systems experience. Based in Paris, she focuses on efficient LLM training, dataset curation, and reproducible evaluation, and has a track record of shipping features that improve TFLOPs logging, gradient accumulation, and resume-from-checkpoint workflows.
7 years of coding experience
1 year of employment as a software developer
Master of Science- MS Big Data & Data Science, Major in Big Data & Data Science, Master of Science- MS Big Data & Data Science, Major in Big Data & Data Science at Ecole Nationale Supérieure des Mines de NancyMasters's degree, MVA (Mathematics, Vision and Learning), Masters's degree, MVA (Mathematics, Vision and Learning) at Ecole normale supérieure Paris-SaclayClasses préparatoires aux grandes écoles, Major in Advanced and Applied mathematics, Computer Science and Physics, Classes préparatoires aux grandes écoles, Major in Advanced and Applied mathematics, Computer Science and Physics at CPGE Moulay Youssef
French, Arabic, English, German
Stackoverflow
1reputation
0reached
1answer
0questions
Github Skills (20)
pytorch10
language-model10
python10
machine-learning10
code-generation10
deep-learning10
trainings10
transformer10
evaluation-framework10
nlp10
modeling10
checkpoint9
acc9
checkpointing9
accelerator9
Programming languages (5)
Github contributions (5)
bigcode-project/bigcode-evaluation-harness
Aug 2022 - Jan 2023
A framework for the evaluation of autoregressive code generation language models.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:1 release, 112 reviews, 242 commits in 4 months
Contributions summary:Loubna contributed to the bigcode-evaluation-harness repository by adding new arguments to the codebase, implementing the use of the `bos_token` as a `eos_token` and refactoring the generation function for code. Additionally, the user was involved in adding references for the MBPP and code-to-text tasks and adding the level_apps argument to the apps metric call. These modifications suggest involvement in the core functionality of the evaluation framework.
nlpcode-generationevaluationmachine-learningrefactoring
huggingface/transformers
Apr 2022 - Aug 2022
🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.
Role in this project:ML Engineer
ML EngineerContributions:26 reviews, 8 commits, 12 PRs in 3 months
Contributions summary:Loubna primarily contributed to the development and enhancement of the CodeParrot training script, a project likely focused on training language models. Their work involved implementing new features for the training process, including TFLOPs logging, gradient accumulation fixes, and checkpointing. They also made code improvements by refactoring, adding functionalities for resuming from checkpoints, and contributing to data preprocessing. Their contributions suggest a focus on optimizing the training pipeline for large language models.
pythonbertspeech-recognitionstate-of-the-artflax
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.