contributor

Toma Iliescu

@toma

Training infrastructure for research-scale workloads. Spends most of his time figuring out why a run that worked on 8 GPUs falls over on 64. Currently working on FSDP recipes and large-scale debugging tooling.

1 article

Training focus

1 article

[2026.10.022]

A research-grade trainer in 400 lines

Most training frameworks are 50,000 lines of code in a trench coat. This is what falls out when you start from FSDP and a will to delete.

Toma Iliescu Training

Oct 09, 2026
25 min