Remote Multilingual Foundation Models Researcher

Translated Srl; Roma; Ibrido; EUR 35.000 - 55.000; Tempo pieno

Vantaggi offerti da questo lavoro

Flexible remote policy

Periodic presence at Rome HQ

Descrizione del lavoro

Translated is seeking an Applied AI Researcher for its Foundation Models team to advance multilingual Large Language Models, pre-training, and data curation at scale. The role spans model design, experiment execution, and benchmarking across languages, with collaboration across researchers and engineers.

The team works on data selection, scaling laws, and distributed training on large GPU clusters, with periodic presence at the Rome HQ and a flexible remote policy.

Competenze

• 3+ years of research/industry experience in Deep Learning, ML, NLP, or Large Language Models.
• Strong understanding of modern deep learning and Transformer-based language models.
• Excellent programming skills in Python.
• Experience designing and running machine learning experiments.
• Familiarity with GPU-based training environments and Unix/Linux systems.
• Ability to analyze experimental results and make research decisions based on empirical evidence.
• Interest in large-scale experimental research and language model pre-training.
• Excellent written and spoken English.
• Ability to work effectively with both researchers and engineers.

Mansioni

• Design and conduct research on Large Language Model pre-training.
• Design experiments, implement them in code, run them at scale, and analyze results.
• Investigate model architectures, optimization strategies, training dynamics, and scaling behavior.
• Research and develop multilingual training strategies.
• Work on data selection, quality, composition, and data mixture experiments for LLM pre-training.
• Evaluate models across multilingual and general-purpose benchmarks.
• Monitor and benchmark the state of the art in Large Language Models.
• Run experiments on large-scale GPU and HPC infrastructure.
• Translate research hypotheses into measurable experiments and actionable decisions for large-scale training.

Conoscenze

Python

Experiment design

Deep Learning

NLP/LLMs

Unix/Linux

GPU-based training

English communication

Collaboration

Strumenti

Megatron Bridge

Descrizione del lavoro

Translated is seeking an Applied AI Researcher for its Foundation Models team to advance multilingual Large Language Models, pre-training, and data curation at scale. The role spans model design, experiment execution, and benchmarking across languages, with collaboration across researchers and engineers.

The team works on data selection, scaling laws, and distributed training on large GPU clusters, with periodic presence at the Rome HQ and a flexible remote policy.