Sciforium

Pre-training Research Engineer

Sciforium
Apply
5 days ago

Base Salary

$165k - $225k/yr

Responsibilities

  • Train large byte-native and multimodal foundation models across massive, heterogeneous corpora.
  • Implement and evaluate model architectures, training objectives, and optimization methods.
  • Develop stable pre-training recipes and conduct scaling experiments for novel architectures.
  • Run ablations and analyze training dynamics, model behavior, and base-model quality.
  • Collaborate with data and distributed training engineers to improve training efficiency, reliability, and scalability.
  • Write robust, performant, production-grade training code and infrastructure.

Requirements

  • 5+ years of experience in machine learning research or engineering, including developing and pre-training large language or multimodal foundation models.
  • Strong general software engineering skills and the ability to write robust, performant training code.
  • Solid understanding of deep learning fundamentals, modern pre-training methods, and related literature.
  • Ability to implement research ideas and evaluate them using baselines, ablations, metrics, and analysis.
  • Hands-on experience running GPU-based training workloads and familiarity with distributed training.
  • MS in Computer Science, Machine Learning, Artificial Intelligence, Mathematics, or a related field.
  • Preferred qualifications include a PhD, extensive JAX/Flax/XLA experience, multi-node pre-training with FSDP, ZeRO, or Megatron, experience developing training recipes and scaling experiments, and ownership of monitored, reproducible training and evaluation pipelines.

Benefits

  • Medical, dental, and vision insurance.
  • 401k plan.
  • Daily lunch, snacks, and beverages.
  • Flexible time off.
  • Competitive salary and equity.

Categories

AI ResearchML Engineering
Sciforium

About Sciforium

11-50 employees
Contact me