Summary

This senior ML engineer role on Nebius's Token Factory team focuses on fine-tuning and inference optimization for large language and multimodal models. You will work with Python, JAX, distributed training, low-precision methods, and large-scale GPU infrastructure; the working arrangement is not specified.

Responsibilities

• Enhance fine-tuning methods for foundation models
• Build model training and evaluation pipelines
• Optimize large language model inference performance
• Investigate low-precision training and inference
• Train large models across multiple computational nodes
• Analyze sharding strategies and hardware performance