Apertus Engineer: Post-training

ETH Zürich; Zürich; Vor Ort; CHF 110.000 - 170.000; Vollzeit

Benefits dieser Stelle

Access to HPC infrastructure

Open-source collaboration

Zusammenfassung

ETH Zürich seeks an engineer to join Apertus post-training efforts, developing, running, and evaluating SFT and RL pipelines to turn base models into capable assistants. The role emphasizes collaboration in an HPC research environment.

You will work across LLM post-training, containerised environments, Slurm jobs on Alps/CSCS, and close coordination with researchers from EPFL and ETH Zürich, contributing to open, multilingual AI models and scalable experiments.

Qualifikationen

• Experience in AI and neural network architectures.
• Hands-on experience with LLM post-training (SFT, preference optimisation) or reinforcement learning.
• Strong collaboration and communication across research and engineering teams.
• Project or study-based experience; formal work experience preferred.
• Flexibility with shifting priorities and schedules.
• Ability to work in an HPC environment with researchers and CSCS engineers.

Aufgaben

• Develop, run, and evaluate scalable post-training workflows for Apertus.
• Build and maintain containerised environments for LLM post-training and RL workloads.
• Adapt containers for execution on Alps/CSCS infrastructure.
• Run Slurm-based training and evaluation jobs.
• Debug distributed execution, checkpointing, filesystem performance, networking, and GPU utilisation.
• Collaborate with researchers and CSCS engineers to improve reliability and performance of large-scale experiments.
• Support SFT, preference optimisation, and reinforcement learning workflows.
• Build RL environments for tasks with verifiable outcomes.
• Run ablation studies and evaluate model behaviour across benchmarks.

Kenntnisse

LLM post-training experience

HPC collaboration

Research collaboration

Distributed training

Software engineering

Ausbildung

MSc or PhD in Computer Science / AI / ML

Exceptional BSc candidates with strong engineering background

Tools

Slurm

Docker / containerization

Megatron-LM / DeepSpeed

veRL / TRL / vLLM

Jobbeschreibung

We are seeking a skilled engineer to join the Apertus post-training effort. The ideal candidate will develop, run, and evaluate the SFT and reinforcement learning pipelines used to turn Apertus base models into capable assistants. This role requires a strong background in LLM post-training, solid software engineering skills, and the ability to work collaboratively in a research-focused HPC environment.

Project background

We train open foundation models with hundreds of billions of parameters on thousands of GPUs on one of the largest AI-ready supercomputers in Europe. The team counts more than a dozen full-time engineers working alongside leading researchers from EPFL and ETH Zürich, has released the Apertus 1 and Apertus 1.5 models, and works with over thirty academic collaborators to deliver fully open (open source), responsibly trained, multilingual, multimodal AI models for research and industry.

Apertus is trained and developed on Alps, the Swiss National Supercomputing Centre's supercomputing infrastructure. The role requires someone who is comfortable working in an HPC environment and collaborating with researchers and infrastructure engineers.

Job description

The engineer will contribute to the development, execution, and evaluation of scalable post-training workflows for Apertus.

Infrastructure and systems engineering

• Build and maintain containerised environments for LLM post-training and RL workloads
• Adapt containers and dependencies for execution on Alps / CSCS infrastructure
• Run and monitor Slurm-based training and evaluation jobs
• Debug failures related to distributed execution, checkpointing, filesystem performance, networking, and GPU utilisation
• Help maintain reproducible training recipes, configuration files, launch scripts, and documentation
• Work with researchers and CSCS engineers to improve the reliability and performance of large-scale experiments

LLM post-training and reinforcement learning

• Support SFT, preference optimisation, and reinforcement learning workflows
• Build and run RL environments for tasks with verifiable outcomes, such as mathematics, code, tool-use, and reasoning
• Implement and run reward modelling, reward calibration, and verifier-based training
• Generate and validate synthetic or gym training tasks
• Run ablation studies comparing algorithms, reward functions, data mixtures, hyperparameters, and infrastructure settings
• Evaluate model behaviour across reasoning, coding, mathematics, instruction-following, multilingual, tool-use, and safety benchmarks
• Debug common post-training issues, including optimisation instability, reward hacking, regressions, and evaluation failures

Profile

Essential

• MSc or PhD in Computer Science, Data Science, Artificial Intelligence, Machine Learning, or a related field
• Exceptional BSc candidates with strong engineering experience will also be considered
• Experience in AI and neural network architectures
• Strong collaboration and communication skills and ability to work across research and engineering teams
• Prior hands-on experience in the core domains of this role is required
• This can be project or study based experience; formal work experience is preferred
• A high degree of flexibility: priorities, tools, and day-to-day tasks shift with training schedules, releases, and a fast-moving field
• Hands-on experience with LLM post-training, be it alignment (SFT, preference optimisation) or reinforcement learning
• This means experience with frameworks such as veRL, slime, Megatron-LM, DeepSpeed, TRL, vLLM, SGLang, or similar tools

Strongly preferred

• Familiarity with distributed training concepts such as data parallelism, tensor parallelism, pipeline parallelism, checkpointing, and GPU communication
• Experience with Slurm or another HPC workload manager
• Experience building or adapting containers for HPC or GPU clusters

Nice to have

• Published research in the domains relevant to this role, or familiarity with recently published research on these topics
• Experience creating verifiable tasks for mathematics, code, reasoning, or tool use
• Familiarity with lower-level GPU/distributed libraries such as NCCL, Transformer Engine, FlashAttention, or communication backends
• Experience with large-scale evaluation pipelines

We offer

• A stimulating academic environment at one of the world's leading technical universities
• The opportunity to work with state-of-the-art supercomputing infrastructure and cutting-edge AI research
• Collaboration with top researchers and engineers from EPFL, ETH Zürich, CSCS, and other Swiss institutions
• Flexible working arrangements, including options for remote work
• Professional development opportunities, including conference attendance and specialised training
• The chance to contribute to open-source projects with global impact
• Access to the broader Swiss academic ecosystem and industry partnerships
• Being part of Switzerland's sovereign AI development, working on technology with national significance
• The role can be based either in Lausanne at EPFL or in Zürich at ETH Zürich

We value diversity and sustainability

In line with our values, ETH Zurich encourages an inclusive culture. We promote equality of opportunity, value diversity and nurture a working and learning environment in which the rights and dignity of all our staff and students are respected. Visit our Equal Opportunities and Diversity website to find out how we ensure a fair and open environment that allows everyone to grow and flourish. Sustainability is a core value for us – we are consistently working towards a climate-neutral future.