Half of my FP8 convolutions were silently running in f32 – a one-line XLA fix
This story is from 2026-09-17. It is preserved in the archive; the latest stories are on the live feed.
A stray return made XLA:GPU skip FP8 rewriting for most of a module I'm training a search-based RL agent (Gumbel MuZero via DeepMind's mctx ) with a small convolutional network, in JAX on a single RTX 5070 (Blackwell, sm_120). The self-play actor is the hot loop: one policy move is a root network e…
Read the full story at DEV Community — Machine Learning ↗
Timeline · 1 report
- 2026-09-17 17:40 · DEV Community — Machine Learning
Half of my FP8 convolutions were silently running in f32 – a one-line XLA fix