Researcher - LLMs

About the role

Huawei Canada is seeking a Researcher to work on large language models (LLMs) within its Noah’s Ark Laboratory in Montreal. This is a 12‑month contract position focused on applied research to enable long‑context LLMs.

Key responsibilities

• Participate in applied research projects focused on enabling long‑context LLMs, including efficient architectures, data curation, and post‑training methods.
• Develop and maintain training, post‑training, and fine‑tuning pipelines, including data preprocessing and evaluation workflows.
• Collaborate with researchers to translate novel ideas into production‑quality implementations.
• Track advancements in LLM architectures, tooling, and systems, and incorporate relevant developments into engineering workflows.
• Publish research in top‑tier machine learning and NLP conferences.

Required profile

• Master’s or Ph.D. in Computer Science, Software Engineering, or a related field, or equivalent practical experience.
• Deep understanding of NLP and large language model fundamentals and state‑of‑the‑art methods, with strong theoretical grounding in machine learning.

Required skills

• Python programming (expert level).
• PyTorch deep‑learning framework.
• Experience with Hugging Face Transformers/Accelerate, Megatron‑LM, vLLM or similar libraries.
• Hands‑on experience with distributed training frameworks such as FSDP, DeepSpeed, DDP.
• Knowledge of modern LLM architectures (e.g., Llama, Qwen) and linear‑attention models (e.g., GDN, Mamba, GLA).