Upload logs/onset_v8/vocal_onset_v8_20260405_085111.log with huggingface_hub

logs/onset_v8/vocal_onset_v8_20260405_085111.log

@@ -0,0 +1,109 @@
+=== onset_v8 — Sun Apr  5 08:51:11 HKT 2026 ===
+SLURM_JOB_ID=128352 node=gpu-4080-410 gpu=0
+v6 MERT cache: 11321 files
+
+=== Phase 1: Train onset_v8 (hidden_dim=128) ===
+Device: cuda  MERT layers: [1, 2, 3, 4]
+Train: 11283  Val: 39
+
+--- MERT feature extraction (11322 tracks) ---
+Loading MERT-v1-95M (frozen)...
+WARNING: feature_extractor_cqt requires the libray 'nnAudio'
+
+/userhome/cs/u3571824/fyp/muse/backend/ml_models/dev/vocal/train_onset_v8.py:446: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
+  scaler = GradScaler(enabled=(args.fp16 and device == 'cuda'))
+/userhome/cs/u3571824/fyp/muse/backend/ml_models/dev/vocal/train_onset_v8.py:328: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
+  with cuda_autocast(enabled=scaler.is_enabled()):
+/userhome/cs/u3571824/anaconda3/envs/bass/lib/python3.10/site-packages/mir_eval/onset.py:49: UserWarning: Reference onsets are empty.
+  warnings.warn("Reference onsets are empty.")
+/userhome/cs/u3571824/anaconda3/envs/bass/lib/python3.10/site-packages/mir_eval/onset.py:51: UserWarning: Estimated onsets are empty.
+  warnings.warn("Estimated onsets are empty.")
+MERT loaded (94M params, frozen)
+  WARN: /userhome/cs/u3571824/fyp/muse/backend/ml_models/dev/vocal/data/processed/audio/mir_st500_162.wav: Calculated padded input size per channel: (1). Kernel size: (2). Kernel size can't be greater than actual input size
+MERT feature extraction: 0 new, 11321 cached, 1 failed
+  WARN: missing cache /userhome/cs/u3571824/fyp/muse/backend/ml_models/dev/vocal/data/processed/audio/mir_st500_162.wav
+OnsetMERTGRU: 0.495M params (MERT frozen)
+
+=== Training onset_v8: 40 epochs ===
+Ep 1/40 (450s) lr=1.00e-03 | train_loss=0.0225 | val_f1=0.3113 P=0.2861 R=0.3732
+  ★ New best: 0.3113
+Ep 2/40 (465s) lr=1.00e-03 | train_loss=0.0212 | val_f1=0.2956 P=0.2698 R=0.3424
+Ep 3/40 (429s) lr=9.98e-04 | train_loss=0.0207 | val_f1=0.4078 P=0.4106 R=0.4204
+  ★ New best: 0.4078
+Ep 4/40 (437s) lr=9.93e-04 | train_loss=0.0203 | val_f1=0.3219 P=0.3057 R=0.3565
+Ep 5/40 (432s) lr=9.85e-04 | train_loss=0.0200 | val_f1=0.4320 P=0.4369 R=0.4334
+  ★ New best: 0.4320
+Ep 6/40 (426s) lr=9.73e-04 | train_loss=0.0197 | val_f1=0.5086 P=0.5320 R=0.4972
+  ★ New best: 0.5086
+Ep 7/40 (417s) lr=9.58e-04 | train_loss=0.0194 | val_f1=0.3959 P=0.4020 R=0.3964
+Ep 8/40 (425s) lr=9.40e-04 | train_loss=0.0192 | val_f1=0.4902 P=0.4959 R=0.4973
+Ep 9/40 (420s) lr=9.19e-04 | train_loss=0.0190 | val_f1=0.4662 P=0.4982 R=0.4431
+Ep 10/40 (429s) lr=8.95e-04 | train_loss=0.0187 | val_f1=0.3356 P=0.3095 R=0.4122
+Ep 11/40 (423s) lr=8.68e-04 | train_loss=0.0184 | val_f1=0.5569 P=0.5493 R=0.6085
+  ★ New best: 0.5569
+Ep 12/40 (422s) lr=8.39e-04 | train_loss=0.0181 | val_f1=0.5431 P=0.5375 R=0.5703
+Ep 13/40 (417s) lr=8.07e-04 | train_loss=0.0179 | val_f1=0.5478 P=0.5595 R=0.5472
+Ep 14/40 (431s) lr=7.73e-04 | train_loss=0.0175 | val_f1=0.5367 P=0.5679 R=0.5168
+Ep 15/40 (419s) lr=7.38e-04 | train_loss=0.0172 | val_f1=0.4847 P=0.4841 R=0.4942
+Ep 16/40 (416s) lr=7.01e-04 | train_loss=0.0168 | val_f1=0.5282 P=0.5415 R=0.5232
+Ep 17/40 (432s) lr=6.62e-04 | train_loss=0.0166 | val_f1=0.5553 P=0.5718 R=0.5533
+Ep 18/40 (432s) lr=6.23e-04 | train_loss=0.0162 | val_f1=0.5408 P=0.5510 R=0.5420
+  Early stopping at epoch 18
+
+Done. Best val_f1=0.5569  ->  /userhome/cs/u3571824/fyp/muse/backend/ml_models/dev/vocal/outputs/onset_v8/best.pt
+
+
+
+
  .../outputs/onset_v8/best.pt:  28%|██▊       |  555kB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  28%|██▊       |  555kB / 1.99MB            
+
+
+
  .../outputs/onset_v8/best.pt:  28%|██▊       |  555kB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  28%|██▊       |  555kB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  28%|██▊       |  555kB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  28%|██▊       |  555kB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  56%|█████▌    | 1.11MB / 1.99MB            
+
+
+
  .../outputs/onset_v8/best.pt:  56%|█████▌    | 1.11MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  56%|█████▌    | 1.11MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  56%|█████▌    | 1.11MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  56%|█████▌    | 1.11MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  56%|█████▌    | 1.11MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  84%|████████▍ | 1.67MB / 1.99MB            
+
+
+
  .../outputs/onset_v8/best.pt:  84%|████████▍ | 1.67MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  84%|████████▍ | 1.67MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  84%|████████▍ | 1.67MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  84%|████████▍ | 1.67MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  84%|████████▍ | 1.67MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt:  84%|████████▍ | 1.67MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt: 100%|██████████| 1.99MB / 1.99MB            
+
+
+
  .../outputs/onset_v8/best.pt: 100%|██████████| 1.99MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt: 100%|██████████| 1.99MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt: 100%|██████████| 1.99MB / 1.99MB            
+
+
  .../outputs/onset_v8/best.pt: 100%|██████████| 1.99MB / 1.99MB            
+No files have been modified since last commit. Skipping to prevent empty commit.
+Phase 1 complete: /userhome/cs/u3571824/fyp/muse/backend/ml_models/dev/vocal/outputs/onset_v8/best.pt