AI Research Engineer, Pre-training LLM & Multi-Modal

Tether develops digital finance products, including stablecoins, digital asset tokenization services, energy solutions for Bitcoin mining, AI and peer-to-peer technology, secure data-sharing applications, and digital learning services.

• Conduct foundational pre-training for LLMs and multimodal models using large distributed servers with multiple nodes and thousands of NVIDIA GPUs
• Design, prototype, and scale architectures, tokenizers, and cross-modal alignment layers
• Source, filter, and curate large-scale textual and multimodal datasets, and establish data pipelines for pre-training
• Execute experiments independently and collaboratively, analyze results, and refine training methodologies
• Investigate, debug, and resolve bottlenecks in model efficiency, computational performance, and multimodal alignment stability during long training runs
• Advance distributed training systems to improve scalability and hardware efficiency on target platforms

• Degree in Computer Science or a related field
• Hands-on experience contributing to large-scale LLM or multimodal pre-training runs on distributed servers equipped with thousands of NVIDIA GPUs
• Familiarity and practical experience with large-scale distributed training frameworks, libraries, and tools
• Deep knowledge of state-of-the-art transformer and non-transformer modifications for improving intelligence, efficiency, and scalability
• Strong expertise in PyTorch and Hugging Face libraries, with practical experience in model development, continual pre-training, and deployment
• Excellent English communication skills
• Будет плюсом: PhD in NLP, Machine Learning, or a related field; a strong AI R&D track record; publications at A* conferences

• Условий в вакансии нет