Commit History

  1. Pre-deadline audit: honest sealed-test numbers in parse hint + model card, lookup timeout, roofline cache fix

    Verified signature
    f3cb4af
  2. Update README.md

    Verified signature
    de0fb20
  3. README: add demo video + social post links (Post it requirement)

    Verified signature
    371d8af
  4. Apple: grey out the VRAM box on Mac (unified memory only)

    Verified signature
    4f270e4
  5. Apple: unified-memory label + chip options without misleading GB

    Verified signature
    8e5e755
  6. Fix: re-parsing a machine without a GPU now clears the previous GPU selection

    Verified signature
    49823c5
  7. spec parser: GPU duration back to 120s (within ZeroGPU cap)

    Verified signature
    ef62a95
  8. Fix spec-parser cold-start hang: pre-cache weights at boot

    Verified signature
    2bb0dc4
  9. Add achievement:fieldnotes tag (blog published)

    Verified signature
    ebd4308
  10. Buy advice: mini-PC category now starts at a $249 Jetson, real cheap-to-big ladder

    Verified signature
    29b7bde
  11. Buy advice: verified, dated prices with live price links per build

    Verified signature
    0348ce3
  12. Fix buy advice: Minimum and Comfortable are now distinct builds

    Verified signature
    de063cc
  13. Polish buy-card 'What to look for' into an iconed spec sheet

    Verified signature
    71ca3cd
  14. Restore line icons + brand-coloured logos (NVIDIA/Windows/Apple etc.)

    Verified signature
    e586a2c
  15. Reskin: CRT-pixel technical-instrument UI (dense option table, roofline centerpiece, pixel icons, phosphor-green, no emoji/em-dash)

    Verified signature
    562dfe1
  16. Audit remediation: XSS fix, pinned loaders, engine bug fixes, honest claims, head_dim KV, hackathon tags

    Verified signature
    412a862

  1. Vision/diffusion compute-roofline speed (FPS + s/image); fine-tune VRAM validation+sweep; flag-gated eager model load

    Verified signature
    44f0592
  2. Laptop/Mobile NVIDIA GPUs with correct VRAM and bandwidth

    Verified signature
    a7a7dde
  3. Honest non-LLM lookups, realistic speeds, clearer gauge, custom/URL coupling, runner icon

    Verified signature
    db17f1f
  4. Full model-URL breakdowns, gauge clarity, CPU honesty, multi-platform buying guide, chart tooltips, runner icon

    Verified signature
    016026a
  5. Catalogue now 110 official models; GGUF size fix

    Verified signature
    c464c37
  6. Interactive results, working priority, fine-tuning mode, honest custom + model lookup

    Verified signature
    0a032c8

  1. Per-position accent borders on the use-case pills

    Verified signature
    7303ce7
  2. GPU spam guard (serialize + button lock), live progress tickers, describe-first layout, real-machine bindings, chart fix for custom VRAM

    Verified signature
    ca2bb8e
  3. Fix spec-parser GPU window; clearer chart colours, unclipped label

    Verified signature
    bca1ea2
  4. Paste box goes live: fine-tuned spec parser fills the form from messy text

    Verified signature
    0e8e243
  5. Load speed predictor via xgboost native format

    Verified signature
    51b6fd5
  6. Speed predictions now come from a model trained on 6.6k real measurements (17.5% median error on unseen hardware), physics formula beyond its measured range

    Verified signature
    0935028
  7. Speed predictions with receipts: bandwidth roofline, real-runs chart, honest provenance

    Verified signature
    ee8ca43

  1. Multi-goal checks (union buying advice), fix vision/audio quality bars, honest niche fallback

    Verified signature
    1bbff15
  2. Add einops for the Mamba hub kernel

    Verified signature
    a87d7a7
  3. Fix kernels version clash: pin to transformers' declared range

    Verified signature
    ebf621f
  4. Real catalogue: 83 verified models, buy-advice mode, live model lookup, license-aware cards

    Verified signature
    e34beb2
  5. Speed up explainer: prebuilt Mamba kernels (transformers 5.8 + kernels hub), tighter answer budget

    Verified signature
    baeaf14
  6. Deploy FitCheck: engine + Nemotron model brick on ZeroGPU

    Verified signature
    869c297
  7. Deploy FitCheck: engine + Nemotron model brick on ZeroGPU

    Verified signature
    12d2e34
  8. Deploy FitCheck UI brick

    Verified signature
    c97ad08
  9. initial commit

    Verified signature
    c2e1718