Handoff: better runtime estimates for planned harness experiments

2026-08-13 · for the agent/session that built the harness and ran the E9 smoke (session de4cf3e4-8580-4249-b330-a7bdb1b6f971, host Battlebook) · repo /home/miro/dev/3dai.iolabs.pointcloud.mlsegmentation

A management presentation is being built with a timescale slide: how long does a 30–50-epoch training run over ~50 corridor segments take? The current numbers are extrapolated from your 3-epoch smoke. You ran the actual training and know the pipeline internals — check the extrapolation below, correct it where you can, and answer the open questions. Write your answers to /tmp/harness-runtime-answers-20260813.md.

What the presentation currently claims (check this)

Anchors measured from your artifacts (runs/E9/version_5 train, version_6 eval, train.log, provenance timestamps):

Extrapolation used in the deck (linear in tile count):

ScenarioEpoch time30 epochs50 epochs
50 segments (~150 tiles), decimated 6.67 cm, local 3080 Ti~15 min~7.5 h~12.5 h
Same on Azure single-node 2×T4 (assumed ×1.5–2 slower)~22–30 min~11–15 h~19–25 h
Full-resolution at 3 cm grid (assumed ×3–5 training points)~1–2 days local~2–5 days T4

Open questions (highest value first)

  1. Linear scaling: is epoch wall time ≈ linear in tile count at batch_size 1, or do dataloader workers / RAM (25 GB local WSL) / caching break that at ~150 tiles? Is a 50-segment local Pointcept run even feasible in 25 GB RAM, or does it force the AML cluster regardless of GPU speed?
  2. Full-res multiplier: once E2 unlocks full-resolution data at a 2–5 cm working voxel, what is the realistic training-point multiplier vs the 6.67 cm decimated smoke data? Is ×3–5 fair at 3 cm?
  3. T4 factor: better estimate than "×1.5–2 vs 3080 Ti Laptop" for spconv SpUNet fp16? Does the single-node 2×T4 job actually use both GPUs, or one?
  4. E1 SPT: expected wall time for the SPT pilot (partition preprocessing + 30–50 epochs) at 50-segment scale on the T4 cluster. Literature anchor in the vault is S3DIS ≈ 3 GPU-h; anything better from the SPT configs/docs you pinned?
  5. Per-epoch validation cost: at 50 segments the val set grows too — does per-epoch val eval stay negligible or start to dominate?
  6. evaluate.py scaling: 71 min for 3 tiles single-core — linear in corridors? Any estimate of the post-vectorization cost (you flagged it as a vectorization candidate alongside the E2 oracle treatment)?

Key artifacts

Constraints