nohup: ignoring input
[eval_v32] Running inference on /home/aiserver/works/fingerprint/sifq/checkpoints/v32/last.pt ...
Loading checkpoint: /home/aiserver/works/fingerprint/sifq/checkpoints/v32/last.pt
Model loaded. Device: cuda
After excluding sensors {'R_500_slap', 'S_500_slap', 'R_1000_slap'}: 16061 records remain
Discovered 16061 records
  4000/16061 done
  8000/16061 done
  12000/16061 done
  16000/16061 done
Done. Scores saved to: /home/aiserver/works/fingerprint/sifq/eval_results/sifq_scores_v32.jsonl
[eval_v32] Inference done: /home/aiserver/works/fingerprint/sifq/eval_results/sifq_scores_v32.jsonl
[eval_v32] Running evaluation (Track 2 + Track 4, exclude: R_1000_slap,R_500_slap,S_500_slap)...
Loaded 16061 SIFQ score records from /home/aiserver/works/fingerprint/sifq/eval_results/sifq_scores_v32.jsonl
After excluding sensors {'S_500_slap', 'R_500_slap', 'R_1000_slap'}: 16061 records remain
[Track 2] Computing sensor invariance...
  Sensor invariance done. Mean KS=0.1375, Pearson=0.7892
[Track 1] Skipped (--skip-track1)
[Track 4] Computing concept grounding (crosstalk matrix)...
  Crosstalk matrix saved.
[eval] Summary table: /home/aiserver/works/fingerprint/sifq/eval_results/v32/eval_summary.txt

============================================================
EVALUATION SUMMARY
============================================================

Track 2 — Sensor Invariance:
  SIFQ      : Mean KS = 0.1375  Pearson = 0.7892

Plots and TXT saved to: /home/aiserver/works/fingerprint/sifq/eval_results/v32
Full summary: /home/aiserver/works/fingerprint/sifq/eval_results/v32/eval_summary.txt
[eval_v32] Eval done. Results in /home/aiserver/works/fingerprint/sifq/eval_results/v32
[eval_v32] Generating score milestone samples...
Saved → /home/aiserver/works/fingerprint/sifq/eval_results/v32/milestone_samples.png
[eval_v32] All done.