LLM Researcher — Long-Context AI & Production Pipelines

Huawei Technologies Canada Co., Ltd.; Montreal (administrative region); On-site; CAD 80,000 - 100,000; Full time

Job summary

A global technology company in Montreal is seeking a Researcher to join its innovative team for a 12-month contract. The role involves participating in cutting-edge research focused on large language models (LLMs) and developing robust training pipelines. Ideal candidates will have a Master's or Ph.D. in Computer Science, strong Python programming skills, and a deep understanding of natural language processing. Proficiency in English is essential for communication with international teams, while a French-speaking environment is prioritized.

Qualifications

• Deep understanding of large language models (LLMs) and state-of-the-art methods.
• Experience with linear-attention models (e.g., GDN, Mamba).
• Hands-on experience with diffusion models is a plus.

Responsibilities

• Participate in applied research on enabling long-context LLMs.
• Develop and maintain training pipelines, including data preprocessing.
• Collaborate with researchers for production-quality implementations.

Skills

Understanding of NLP

Programming skills in Python

Expertise in PyTorch

Experience with LLM training and inference

Strong communication skills

Education

Master’s or Ph.D. in Computer Science

Tools

Hugging Face Transformers

Megatron-LM

Distributed training systems (FSDP, DeepSpeed)

Job description

A global technology company in Montreal is seeking a Researcher to join its innovative team for a 12-month contract. The role involves participating in cutting-edge research focused on large language models (LLMs) and developing robust training pipelines. Ideal candidates will have a Master's or Ph.D. in Computer Science, strong Python programming skills, and a deep understanding of natural language processing. Proficiency in English is essential for communication with international teams, while a French-speaking environment is prioritized.