Twitxy Saved

Manish Tiwari

@compmanish

AI Engineer | I build the plumbing LLM apps run on: sandboxes, guardrails, inference benchmarks. 41 merged OSS PRs, saferun-ai on PyPI. Open to remote worldwide
273 Following 63 Followers 2K Posts

Latest posts 5

Pinned post

Manish Tiwari @compmanish
Every "we have 600 GB of pooled VRAM" spreadsheet is fiction.

A model has to fit on ONE node before it can train on any of them. So the ceiling is your largest single GPU, not the sum.

I built a planner to compute this properly. Here's what it says. 馃У
Manish Tiwari @compmanish
Finally a rare addition into the syllabus of Traditional ML!!
Swann Bessa Verified account @swann_bessa 4d
Introducing e-means, an algorithm beating k-means in 10x fewer epochs.

With order of magnitude wall-clock speedups, e-means is released in FAISS and needs no hyperparameter tuning beyond k. It is robust to various scales and datasets. Still skeptical ? Just try e-means [1/10]馃У
View quoted post on X