Lead machine learning engineer Model Architectures ampCapabilities Team - London, United Kingdom - Nebius

Description

Thecompany

Nebius is a moderntechnology venture offering strategic partnerships to leadingcompanies around the world empowering them to create their ownlocal hyperscaler platforms and become trustworthy providers ofcloud services and technologies in their own regions. In additionto innovative software and hardware including server racks designedinhouse Nebius provides a launchready business model andcustomizable tools for support sales andmarketing.

Our aim is to empower our partners tocreate their own IT infrastructure and deliver cuttingedgedisruptive cloud solutions to local markets while keeping securityand compliance with international standards like ISO and GDPR toppriorities.

Were a global company with offices inthe Netherlands Israel andSerbia.

Ourteam

Nebius was founded by a coreteam of engineers and business professionals with a proven trackrecord of using cloud technologies to create value for otherbusinesses. We know from experience that cuttingedge technologiescan only make an impact if their innovation is matched by the levelof the experts managing them so as Nebius expands our core priorityis to attract the most qualified enthusiastic and drivenindividuals we can to join our growingteam.

Nebius Large Language Models(LLM) team is dedicated to pushing the boundariesoflanguage modellingtechnology.We are focused on developing astateoftheart LLM technological stack that spanswebscaledata collection foundational modeltraining and alignment. Our overarching objective istopioneer cuttingedge language generationtechnology for both internal use andcustomerapplications driving the evolutionof the next generation of AIpoweredproducts.

Therole

We are currentlyin search of the team lead for the Model Architectures &Capabilities team.The team is responsiblefor pushing forward the capabilities of the models small andlargethat we train inhouse. This includesfinding model architectures that efficiently achievethedesired capabilities scaling thesemodels to the limits of our hardware and exploringnovelideas that could potentially expandof what ispossible.

Inthis position your responsibility will beto:

• Leadthe team responsible for model architectures andcapabilities
• Definestrategy and tactics i.e. figure out what research and engineeringdirections topursue to push the technologyforwards and help the team plan and executetheexperiments that will get us thereefficiently
• Ensurehigh standards of engineering and research activities within theteam
• Keepimproving the design our internal infrastructure for training largemodels toensure it keeps being fast andflexible despite the technology movingforwards
• Mentorour engineers andresearchers

Weexpect you tohave:

• Aprofound understanding of theoretical foundations of machinelearning
• Deepexpertise in modern deep learning for language processing andgeneration
• Substantialexperience with pretraining large models on hugeclusters
• Goodunderstanding of performance aspects of large neural networktraining (shardingstrategies customkernels hardware featuresetc.)
• Strongsoftware engineering skills (we mostly usepython)
• Deepexperience with modern deep learning frameworks (we usejax)
• Proficiencyin contemporary software engineering approaches includingCI/CDversion control and unittesting
• Strongcommunication and leadershipabilities

Itwould be an added bonus if youhad:

• Bachelorsdegree in Computer Science Artificial Intelligence Data Science orarelated field. Masters or PhDpreferred
• Trackrecord of building and delivering products (not necessarilyMLrelated) in adynamic startuplikeenvironment
• Experiencein engineering complex systems such as large distributed dataprocessingsystems or highload webservices
• Opensourceprojects that showcase your engineeringprowess
• Excellentcommand of the English language alongside superior writingarticulationand communicationskills
  

Doesall that sound like your kind of challenge Then joinus