Vicki Boykis on ONNX inference for retrieval and ranking: slides from DataPhilly talk
vboykis · x · 2026-10-30
Data engineer Vicki Boykis gave a DataPhilly talk on inference in production, with slides now public. Topics: embedding-based retrieval and ranking, ML model artifact creation and inference, general vs. smaller models, and latency/ONNX conversion tradeoffs.
Q&A covered what can be embedded, self-hosting vs. API inference tradeoffs, and whether a model's data or architecture matters more. She notes her first DataPhilly talk in 2013 was about data access frustrations—a problem she spent the next decade solving as a developer.
More from Research
- Eilenberg's visionary 'Automata, Languages and Machines' gets a fresh recommendation — abhi9u · 2026-10-30
- Learning RL from scratch: one-page notes site de-jargonizes reinforcement learning basics — East-Muffin-6472 · 2026-10-30
- "Your LLM evaluation is lying to you": temp-0 Qwen3-235B gives 80 different answers in 1000 runs — irombie · 2026-10-30
- inclusionAI open-sources ZwZ2 vision models: +4.99 points across 25 benchmarks without zoom tools — jacek2023 · 2026-10-30
- Epoch data: leading chips get ~40% more efficient yearly, but AI power draw keeps climbing — teortaxesTex · 2026-10-30
- Soft-MTP: a one-line fix removes noisy gradients from multi-token prediction — pmttyji · 2026-10-30