Post-Training Engineer - Apertus — EPFL
CHF 49'500 - 75'000
EPFL · Lausanne (VD)
- Location
- Lausanne
- Contract
- full-time
- Posted
- 21 days ago
SalaryCHF 49'500 - 75'000
Role overview
Introduction The Apertus project, a joint effort between EPFL and ETH Zürich, is seeking a practical and motivated engineer to help build the next generation of open foundation models.
The successful candidate will help develop and run post-training and reinforcement learning pipelines for the Apertus project.
Apertus is trained and developed on Alps, the Swiss National Supercomputing Centre’s supercomputing infrastructure.
- Introduction The Apertus project, a joint effort between EPFL and ETH Zürich, is seeking a practical and motivated engineer to help build the next generation of open foundation models.
- The successful candidate will help develop and run post-training and reinforcement learning pipelines for the Apertus project.
- Main duties and responsibilities The engineer will contribute to the development, execution, and evaluation of scalable post-training workflows for Apertus. Infrastructure and systems engineering
- Build and maintain containerized environments for LLM post-training and RL workloads.
Main responsibilities
- Main duties and responsibilities The engineer will contribute to the development, execution, and evaluation of scalable post-training workflows for Apertus. Infrastructure and systems engineering
- Build and maintain containerized 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 utilization.
- 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 optimization, and reinforcement learning workflows.
- Build and run RL environments for tasks with verifiable outcomes, such as mathematics, code, tool-use, and reasoning.
Additional details
- Main duties and responsibilities The engineer will contribute to the development, execution, and evaluation of scalable post-training workflows for Apertus. Infrastructure and systems engineering
- Debug common post-training issues, including optimization instability, reward hacking, regressions, and evaluation failures.
- Strong collaboration and communication skills and ability to work across research and engineering teams. Strongly preferred
- Experience with frameworks such as veRL, slime, Megatron-LM, DeepSpeed, TRL, vLLM, SGLang, or similar tools.
- Experience with large-scale evaluation pipelines.
Notes and original content
- Main duties and responsibilities The engineer will contribute to the development, execution, and evaluation of scalable post-training workflows for Apertus.
- Infrastructure and systems engineering
- Profile Essential
- Strong collaboration and communication skills and ability to work across research and engineering teams.
- Strongly preferred
- Nice to have
