Writing
Articles
Long-form write-ups of the work. For shorter technical notes, see Notes.
- Python projects are twice as likely as Java ones to leave most outside pull requests unanswered ↗23 Oct 2026
Hashnode · about Porchlamp
- A hot-expert cache made a 181 GiB model faster on one 96 GiB GPU ↗22 Oct 2026
DEV · about GLM-5.3-Flash on one GPU
- Why a Keras model fails to convert to TFLite: 11 known cases, 5 traps ↗22 Oct 2026
DEV · Also on Medium · about Keras on an Android phone
- Keras to TFLite on Android: wrong normalization and a backwards top-5 ↗22 Oct 2026
DEV · Also on Medium · about Keras on an Android phone
- Two Android ML tests that could never fail, and a TFLite check that can ↗22 Oct 2026
DEV · about Keras on an Android phone
- Keras #22561 speedups: compiled Keras on PyTorch now generates text 11.6x faster ↗16 Oct 2026
DEV · about Keras on PyTorch: compiled generation 11.6x faster
- Exporting Keras models to LiteRT ↗21 Jul 2026
The official keras.io guide to LiteRT export. · keras.io · about One call from a Keras model to a phone: LiteRT export