LLM / ML Engineer

We are looking for an LLM / ML Engineer with practical experience in training language models.

The task is to train specialized models with 1B–10B parameters from scratch. Main areas: software development, understanding complex Russian texts, and function calling.

Projects for clients under NDA — we will discuss details in person.

Tasks: – Prepare data and build training pipelines. – Conduct experiments with architectures, methods, and training modes. – Train models from scratch and adapt them to target tasks. – Evaluate model quality, analyze errors, and improve results.

Expectations: – Practical experience in training language models: from data preparation to result evaluation. – Experience with similar projects or adapting pre-trained models using methods beyond SFT. – Proficient in Python and PyTorch. – Understanding of transformer training principles and experience with GPUs.

Bonus points: experience with distributed training, memory and training speed optimization, working with models for code and tool invocation.

It would be great if you could briefly describe your projects and your role in them.