Haibin Lin leads LLM systems at Bytedance Seed, where he optimizes training infrastructure for foundation models across more than 10, 000 GPUs. A Carnegie Mellon graduate, his prior experience includes developing Apache MXNet at Amazon. He focuses on enhancing the efficiency of both pre-training and post-training processes.
His work on the "MegaScale" paper details the complexities of scaling large language model training to over 10, 000 GPUs.
Read the full overview →Behavioral traits and communication patterns that shape how this person evaluates opportunities, builds trust, and makes decisions in professional settings.
Dominance, influence, steadiness, and calculativeness scores with guidance on how this person prefers to communicate and decide.
Explore more public profiles from related professionals.
Discover additional public profiles from our index.
Looking for someone else? Search here for anyone.