ZipRecruiter

Log In

1

Deepspeed Jobs in California (NOW HIRING)

Hedra

Research Engineer

San Francisco, CA · On-site

$175K - $275K/yr

Build distributed training infrastructure using PyTorch, FSDP, and DeepSpeed * Work with multimodal data pipelines involving video, sensory inputs, and action sequences * Evaluate model performance ...

next page

Showing results 1-20

All JobsDeepspeed JobsDeepspeed Jobs in California

Deepspeed information

What are some common challenges faced by engineers working with DeepSpeed and how can they be addressed?

Engineers working with DeepSpeed often encounter challenges related to optimizing large-scale model training, such as managing memory efficiency and tuning distributed training parameters. Troubleshooting issues like gradient accumulation, parallelism strategies, and ensuring compatibility with different hardware setups can be complex. Collaborating closely with data scientists, DevOps, and research teams is essential for addressing these challenges, as is staying updated with the latest DeepSpeed releases and documentation. Regular participation in code reviews and knowledge-sharing sessions can also help engineers overcome technical hurdles and continuously improve model performance.

What is Deepspeed?

Deepspeed is an open-source deep learning optimization library developed by Microsoft, designed to enable distributed training of large-scale models efficiently. It helps researchers and engineers train models that are too large to fit in the memory of a single GPU by offering features like ZeRO optimization, mixed-precision training, and advanced parallelism techniques. Deepspeed is widely used in the machine learning community for its scalability and performance improvements, making it easier to train state-of-the-art models on vast datasets. The library integrates seamlessly with PyTorch and supports training on multiple GPUs and even across multiple machines.

What is the difference between Deepspeed vs Data Scientist?

AspectDeepspeedData Scientist
Required credentialsKnowledge of machine learning frameworks, programming skills in Python, experience with AI model trainingDegree in Data Science, Statistics, Computer Science, or related fields; strong analytical skills
Work environmentAI research labs, tech companies, cloud computing environmentsBusiness, tech companies, research institutions
Industry usageAI model training, deep learning optimizationData analysis, predictive modeling, business insights

Deepspeed focuses on optimizing large-scale AI model training and deep learning performance, while Data Scientists analyze data to generate insights and build predictive models. Both roles require technical skills but serve different purposes within the AI and data ecosystem.

What are the key skills and qualifications needed to thrive as a DeepSpeed Engineer, and why are they important?

To thrive as a DeepSpeed Engineer, you need a solid background in machine learning, deep learning frameworks (such as PyTorch), and distributed systems, often supported by a degree in computer science or a related field. Proficiency with DeepSpeed, parallel computing libraries, and cloud platforms, along with familiarity with tools like CUDA and NCCL, is typically expected. Strong problem-solving abilities, collaboration, and adaptability are crucial soft skills for optimizing large-scale AI models and working with cross-functional teams. Mastering these skills ensures efficient development and deployment of high-performance, scalable AI solutions in demanding environments.
What are popular job titles related to Deepspeed jobs in California? For Deepspeed jobs in California, the most frequently searched job titles are:
What job categories do people searching Deepspeed jobs in California look for? The top searched job categories for Deepspeed jobs in California are:
What cities in California are hiring for Deepspeed jobs? Cities in California with the most Deepspeed job openings:
Deepspeed jobs near you

Machine Learning Infrastructure Engineer

MBZUAI (Mohamed bin Zayed University of Artificial Intelligence)

Sunnyvale, CA • On-site

$125K - $164K/yr

Full-time

Re-posted 13 days ago

Job description

Job Summary:
MBZUAI is a dedicated research lab focused on advancing research and nurturing the next generation of AI builders. The role involves extending and scaling training systems for cutting-edge foundation model development, working alongside world-class researchers and engineers.
Responsibilities:
• Extend distributed training frameworks (e.g., DeepSpeed, FSDP, FairScale, Horovod)
• Implement distributed optimizers from mathematical specs
• Build robust config + launch systems across multi-node, multi-GPU clusters
• Own experiment tracking, metrics logging, and job monitoring for external visibility
• Improve training system reliability, maintainability, and performance
• Extend or modify training frameworks (e.g., DeepSpeed, FSDP) to support new use cases and architectures.
• Translate mathematical optimizer specs into distributed implementations.
• Create and debug multi-node launch scripts with flexible batch sizes, parallelism strategies, and hardware targets.
• Build systems for experiment tracking, job monitoring, and logging usable by collaborators and researchers.
• Write production-quality code and tests for ML infra in PyTorch or JAX; ensure reliability and maintainability at scale.
Qualifications:
Required:
• 5+ years of experience in ML systems, infra, or distributed training
• Experience modifying distributed ML frameworks (e.g., DeepSpeed, FSDP, FairScale, Horovod)
• Strong software engineering fundamentals (Python, systems design, testing)
• Proven multi-node experience (e.g., Slurm, Kubernetes, Ray) and debugging skills (e.g., NCCL/GLOO)
• Ability to implement algorithms across GPUs/nodes based on mathematical specs
• Experience working on an ML platform/ infrastructure, and/or distributed inference optimization team
• Experience with large-scale machine learning workloads (strong ML fundamentals)
Preferred:
• Exposure to mixed-precision training (e.g., bf16, fp8) with accuracy validation
• Familiarity with performance profiling, kernel fusion, or memory optimization
• Open-source contributions or published research (MLSys, ICML, NeurIPS)
• CUDA or Triton kernel experience
• Experience with large-scale pre-training
• Experience building custom training pipelines at scale and modifying them for custom needs
• Deep familiarity with training infrastructure and performance tuning
Company:
Official account of Mohamed bin Zayed University of Artificial Intelligence. Dedicated to research, innovation, and empowering brilliant minds in AI. Founded in 2019, the company is headquartered in Abu Dhabi, ARE, with a team of 51-200 employees. The company is currently Growth Stage.

Apply