spec parser: GPU duration back to 120s (within ZeroGPU cap)

Fix spec-parser cold-start hang: pre-cache weights at boot

Add achievement:fieldnotes tag (blog published)

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

Buy advice: verified, dated prices with live price links per build

Fix buy advice: Minimum and Comfortable are now distinct builds

Polish buy-card 'What to look for' into an iconed spec sheet

Restore line icons + brand-coloured logos (NVIDIA/Windows/Apple etc.)

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

Audit remediation: XSS fix, pinned loaders, engine bug fixes, honest claims, head_dim KV, hackathon tags

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

Laptop/Mobile NVIDIA GPUs with correct VRAM and bandwidth

Honest non-LLM lookups, realistic speeds, clearer gauge, custom/URL coupling, runner icon

Full model-URL breakdowns, gauge clarity, CPU honesty, multi-platform buying guide, chart tooltips, runner icon

Catalogue now 110 official models; GGUF size fix

Interactive results, working priority, fine-tuning mode, honest custom + model lookup

Per-position accent borders on the use-case pills

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

Fix spec-parser GPU window; clearer chart colours, unclipped label

Paste box goes live: fine-tuned spec parser fills the form from messy text

Load speed predictor via xgboost native format

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

Speed predictions with receipts: bandwidth roofline, real-runs chart, honest provenance

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

Add einops for the Mamba hub kernel

Fix kernels version clash: pin to transformers' declared range

Real catalogue: 83 verified models, buy-advice mode, live model lookup, license-aware cards

Speed up explainer: prebuilt Mamba kernels (transformers 5.8 + kernels hub), tighter answer budget

Deploy FitCheck: engine + Nemotron model brick on ZeroGPU

Deploy FitCheck: engine + Nemotron model brick on ZeroGPU

Deploy FitCheck UI brick

initial commit