J
Staff Research Engineer (Pre-training)
JetBrains
Contract type
Ongoing
Work mode
On-site · On-site (see listing for address)
Experience
Lead / principal · Not stated
Level inferred from job title
Job description
Key details
- Work with stakeholders to convert business requirements into technical specifications
- Train LLMs from scratch on a large GPU cluster
- Collect and process pre-training and fine-tuning datasets
- Support and improve existing subsystems
- Company mission
- We create an open and inclusive workplace where great ideas can come from anyone, anywhere
Primary stack
Core technologies
Python (Programming Language)Kubernetes
Benefits
- Information not specified
Requirements & details
- Experience in design, deployment, and support of production ML systems
- Strong theoretical background in NLP and transformer-based approaches
- Proficiency with modern deep learning frameworks such as PyTorch and common libraries for NLP
- Experience in distributed training of multi-billion parameter models
- Attention to detail and great communication skills
- Experience with LLM inference frameworks such as vLLM, DeepSpeed, TensorRT is a plus
- Experience with LLM alignment techniques such as RLHF/RLAIF is a plus
- Experience with MLOps tools and practices, including CI/CD for ML, is a plus
- Experience with K8s and Kubeflow is a plus
- Scientific publications in the NLP field are a plus
- Python, PyTorch, HuggingFace, Kubeflow, Weights & Biases, Git, TeamCity, vLLM, DeepSpeed, TensorRT, K8s
- Python (Programming Language)
- Kubernetes