Latest Interviews
Showing 1–2 of 2 interview transcripts.
Clear all filters- Jane Street1h 6m
The Uncertain Art of Accelerating ML Models with Sylvain Gugger
Sylvain Gugger, Ron Minsky, Jeremy Howard, Mark Mandelmann, Mark Mirchandani, Francesc Campoy, Gabriel Sanchez
Former fast.ai co-author Jeremy Howard discusses his transition from mathematics education to optimizing machine learning infrastructure at Jane Street, highlighting breakthroughs in learning rate schedules and image resizing that previously secured top benchmark placements. He details the development of the Hugging Face Accelerate library, a lightweight tool designed to abstract complex hardware parallelism and eliminate boilerplate code for training across diverse GPUs and TPUs. The discussion further explores the architectural constraints of financial data, the dominance of PyTorch's iterative execution model, and Jane Street's rigorous approach to reproducibility and custom model development for high-frequency trading.
- Jane Street1h 23m
Memory Management with Stephen Dolan
Stephen Dolan, KC Sivaramakrishnan, Ron Minsky, Mark Mandelmann, Simon Hughes, Matt Sullivan, Mark Blyther, Paul Lewisohn, Simon Verstraeteker
Stephen Dolan discusses advanced optimizations in the OCaml language, detailing how generational, incremental garbage collection and software pre-fetching reduce memory access latency by up to 90%. He introduces local types and unboxed representations to enable safe stack allocation and native memory layouts, thereby eliminating heap overhead for temporary values while maintaining automatic memory safety without manual lifetime annotations. The presentation concludes with a roadmap for integrating multi-core garbage collection and lifting GC barriers to support denser data structures, balancing performance with ergonomic simplicity.